<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Seth Snyder on Medium]]></title>
        <description><![CDATA[Stories by Seth Snyder on Medium]]></description>
        <link>https://medium.com/@sethdanielsnyder?source=rss-7955690dd2ca------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*dx6PJo6zWQ5GHWtmGB_gqw.jpeg</url>
            <title>Stories by Seth Snyder on Medium</title>
            <link>https://medium.com/@sethdanielsnyder?source=rss-7955690dd2ca------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 06:46:31 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@sethdanielsnyder/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[What does it mean to be a truly mission-driven organization?]]></title>
            <link>https://medium.com/@sethdanielsnyder/what-does-it-mean-to-be-a-truly-mission-driven-organization-28e5fe8f44f9?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/28e5fe8f44f9</guid>
            <category><![CDATA[mission-statement]]></category>
            <category><![CDATA[healthcare-technology]]></category>
            <category><![CDATA[healthcare-design]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[healthtech]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Wed, 30 Nov 2022 13:01:58 GMT</pubDate>
            <atom:updated>2022-11-30T13:01:58.722Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qd7muoCwC9BkzVQa8jrQSA.png" /></figure><p>As anyone who’s been a part of building an early-stage startup knows, hitting key business milestones like proving product-market fit and acquiring users is crucial to securing the funding you need to continue growing. It’s all too easy to get into the groove of building the next thing and the next thing, faster and faster, that before you know it, nobody’s talking about the mission anymore. So where, in all of that swirl and the late nights building new technology and the rapid-growth hiring, do you fit in the mission? How do you make sure the team, as it scales, remains singularly focused on the mission that pulled you all together in the first place?</p><p>First, we have to understand what it means to be a truly mission-driven organization. After casually interviewing dozens of my <a href="https://coherehealth.com/">Cohere Health</a> colleagues, I synthesized their conception of what mission-driven means into three key themes:</p><ol><li><strong>Belief:</strong> Everyone in the organization shows up to work every day inspired and motivated to do their best because of their belief in the mission.</li><li><strong>Impact:</strong> Everyone in the organization is motivated more by our ability to make a positive impact on society than by making money.</li><li><strong>Unity:</strong> Everyone in the organization is collaborating to reach the same goal, and they understand how their day-to-day work supports that goal.</li></ol><p>Here’s how we’ve thought about these three themes throughout our time together at Cohere:</p><figure><img alt="Belief" src="https://cdn-images-1.medium.com/max/1024/0*Mhx8zVf8Q5ukNOk2" /></figure><p>We knew we had an inspiring mission for folks to believe in — to tackle the wicked problems of a healthcare system that leaves patients lost and confused. We envision a world in which a patient, their providers, and their health plan are aligned on an optimal care path — one that drives the best possible outcomes at the lowest possible cost, with the least possible administrative effort.</p><p>This focus and the results we were able to deliver helped us to expand our contracts with initial customers and scale from a founding set of 8 employees to a nationally distributed team of over 500 in just over 2 years. It has helped guide our development of cutting-edge machine learning technologies to automate archaic workflows. It has inspired us to build software and service experiences that demonstrably solve issues for both our users and our customers.</p><p>As we grew at such a rapid pace, we knew we needed to ensure we brought on people who were not only talented, intelligent, and capable, but who also had a passion for tackling wicked problems like ours. We screened for folks who could articulate why they thought Cohere’s work was meaningful and impactful; why it mattered to them personally. Oftentimes, candidates would tell a personal story about what moved them to come to healthcare. For example, a family member getting sick opened their eyes to the injustices and lack of guidance in the healthcare system. Prior experience in healthcare was not required, however, so long as candidates could speak to what now drew them to healthcare and how the missions of their past experiences motivated them.</p><p>Once folks are onboard, we find ways to remind everyone of the mission and the deep-seated problems with the healthcare system we’re working to improve. We share real patient stories at All Hands meetings to ensure we stay grounded in the people at the center of the reason to believe in our mission. We share and discuss our own personal experiences with the painful prior authorization process or other barriers to care we or our families encounter. Our UX Research team shares stories they hear from providers about the frustrations they face in the process of delivering care to their patients. And we frequently invite the chairman of our board, Dr. Gary Gottlieb, to speak to us about his experiences leading and transforming health systems in Boston and in the poorest countries around the world. Our teams leave these presentations and discussions freshly inspired and motivated, and their belief in our mission reignited.</p><figure><img alt="Impact" src="https://cdn-images-1.medium.com/max/1024/0*U5eHCjRoxfN9xnCT" /></figure><p>Incorporating the mission into the work we do day to day can be a challenge given that our entry point in patient journeys is indirect. The application that we make available today is designed to capture clinical information so as to make historically manual decisions automatic, thereby decreasing the time folks need to spend on these tasks. What this means is that our objective is to design technology that allows users to spend time on other, more valuable tasks as opposed to prior authorization or clinical review. As a result, sometimes the projects that teams are working on can get so technical and detailed around the administrative processes of utilization management that it can be hard to draw a line connecting the work and the broader mission around improving clinical care and patient experience. The pressures of scale and growth also mean that we can easily get swept into the rhythm of shipping new features again and again and forget to look up and remember the reason we’re all here working so hard. To address this, we task team leads and managers with the responsibility of weaving the mission into projects in a few ways:</p><ul><li>A checklist of measurable mission-driven outcomes is written into every one of our initiative documents so that as teams kick off a project, they know exactly which aspects of the mission this project will contribute to.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*u4RQQXcKxQWcvraY" /></figure><ul><li>Through a collaborative process of refining the initiative documentation, the team aligns on how to measure the actual impact the initiative will hopefully have on the mission so we can be sure to have a clear way to tie it back.</li><li>Weekly initiative updates get posted on Slack which include the impact the initiative is intended to have on the mission as a reminder of where we’re headed and why we’re doing it. Teams report that these updates drive home how the work they’re doing day to day contributes to improving the healthcare system.</li><li>As results are reported on the success of the initiative, KPI impact is framed in the context of the mission.</li><li>Quarterly, our product, design, engineering, clinical, operational, and marketing teams gather together in our Boston HQ to celebrate wins from the past quarter and align on our goals and plans for the upcoming quarter. The event is always an opportunity for company leaders to not only show what we shipped but also how this work helps us move towards the mission we all signed up for when we joined the company.</li></ul><figure><img alt="Unity" src="https://cdn-images-1.medium.com/max/1024/0*J952B_xtE0ecsX0F" /></figure><p>As we grew from a team of a few dozen — when everyone knew everyone — to a team of many teams in the hundreds, it became harder to ensure everyone was working to achieve the same outcomes. Furthermore, healthcare is a sector rich with misaligned incentives and it can often be challenging or uncomfortable to disrupt the status quo en route to achieving mission-driven company goals. We knew that in order to be successful, we’d need to unify around our continued fight for the mission even when the work met resistance or uncomfortable decisions needed to be made. We’d need to continuously demonstrate that the company as a whole was acting in ways that aligned with the mission.</p><p>To ensure every team member understood how their individual and team goals ladder up to the company’s goals and that the company’s goals aligned with the mission, we instituted a quarterly alignment ritual:</p><ol><li>Focus areas: The process starts with our executive team setting an annual strategy comprised of a small set of focus areas and then sharing updates to the strategy throughout the year as needed.</li><li>Initiatives: Then a group of key VPs and Directors we call the Operating Team develop a quarterly priority list of every company initiative they think we need to take on to achieve the company strategy. Each initiative has a link to a focus area and a set of key success metrics. The executive team reviews and approves that list and the work begins.</li><li>Sprints: As the product, design, engineering, and/or clinical work gets underway in two-week sprints, initiative leads are responsible for documenting and sharing key updates on the initiatives. When there are initiatives at risk, those get elevated to discussions with the Operating Team and exec team if needed.</li><li>Daily progress: Teams make progress towards initiatives with full visibility into the impact. Frequent progress updates are also shared on Slack in public channels so everyone in the company has an equal ability to follow along. After an initiative launches and the data has been collected, managers report out on the key success metrics.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*BbpGMBnVdSXtbv4w" /></figure><p>Since instituting this process company-wide, teams have reported that they have a stronger understanding of Cohere’s strategy, better clarity around priorities, and improved visibility into cross-functional work that might affect their team. And this sense of unity has been a crucial aspect of not only feeling like we’re a mission-driven organization, but also being able to measure it, too.</p><figure><img alt="Conclusion" src="https://cdn-images-1.medium.com/max/1024/0*Phr57R-SOWKX2v94" /></figure><p>So if you find yourself in a quote-unquote “mission-driven organization” and you feel the mission is getting lost in the shuffle, try some of these tactics. I’d love to hear what works for you!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=28e5fe8f44f9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to land a product design job]]></title>
            <link>https://medium.com/design-bootcamp/how-to-land-a-product-design-job-ff08583dcdb7?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/ff08583dcdb7</guid>
            <category><![CDATA[portfolio]]></category>
            <category><![CDATA[hiring]]></category>
            <category><![CDATA[ux-design]]></category>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Tue, 17 May 2022 19:53:14 GMT</pubDate>
            <atom:updated>2022-05-20T04:56:20.526Z</atom:updated>
            <content:encoded><![CDATA[<h4>The truth behind what design team hiring managers look for throughout the hiring process</h4><figure><img alt="Looking for a job" src="https://cdn-images-1.medium.com/max/1024/1*AoR-2ZvQFUQTq_rhKrLEhw.png" /></figure><p>As of the publishing of this, I’ve reviewed 856 candidate applications during my two-year-long hunt for talented designers to join our team at <a href="https://coherehealth.com/">Cohere Health</a>. Out of those, I’ve hired 9 amazing people who are truly the most thoughtful and talented designers I’ve ever had the pleasure of working with. While 9 out of 832 isn’t a great ratio, we have had a 100% correlation between the ratings our interviewers gave candidates in the interview process and those candidates’ eventual performance ratings, which <a href="https://joulee.medium.com/">Julie Zhou</a> says in <a href="https://www.juliezhuo.com/book/manager.html">The Making of a Manager</a> is exceedingly unusual. In other words, our fine-tuned interview process gives us an exceptionally high degree of confidence that candidates who make it all the way through will thrive. So what made these 9 designers such a good fit for us? What made them stand out in the first place? How did they shine in each stage of our interview process?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FmNheZjoqd5t0qjxNRb%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Fgodaddy-mNheZjoqd5t0qjxNRb&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2FmNheZjoqd5t0qjxNRb%2Fgiphy.gif%3Fcid%3D790b7611bc73af9345af8e62680f3450dbc529a32b96215c%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="283" frameborder="0" scrolling="no"><a href="https://medium.com/media/9bd9b3a61b37b8f30325bba5964cabd1/href">https://medium.com/media/9bd9b3a61b37b8f30325bba5964cabd1/href</a></iframe><p>Let’s start with the application, which is really all about the resume and portfolio (I’d like to say I read everyone’s cover letter, but I don’t).</p><h3><strong>Resume</strong></h3><p>I start by scanning for your most recent job experience, hunting for clues that you have had the professional opportunities to learn the skills that will make you a good fit for the role. Do you have experience doing product design at a startup or tech company where you would have collaborated with product managers and engineers? Do you have experience working in healthcare or a similarly complex industry? Cohere is a healthtech startup, so look for opportunities to customize your resume to highlight those things I’m going to be looking for. This will increase the likelihood of me moving onto your portfolio. Same goes for any role at any company — think about what the hiring manager is going to be looking for in your resume and make it easier for us to find it.</p><p>Please pay close attention to the number of years of professional experience the roles require — we do not artificially inflate those expectations. While I’m open to the possibility of exceptions to this, there really isn’t a replacement for real world experience building, shipping, and iterating on products in a professional setting like a startup or big tech company. That said, if you feel strongly that you have developed the design skills, product sense, and soft skills that come from professional experience in other ways, we’d love to hear about it. And don’t worry — we (and most companies like us) are continuously growing so keep an eye out for a role that better matches your experience.</p><p>While I’m not expecting your resume to be a graphic design masterpiece, submitting merely simple text or plain word docs will not help your cause. Show me that you used a design tool to design your resume. Making it legible and easy to scan shows me that you care about craft and are detail-oriented with everything you make.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2Fl2R098V90nJfuHKg0%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Ftv-land-younger-youngertv-l2R098V90nJfuHKg0&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fl2R098V90nJfuHKg0%2Fgiphy.gif%3Fcid%3D790b7611248523f13df0cd006614a79f4c41ec662355658b%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="245" frameborder="0" scrolling="no"><a href="https://medium.com/media/6abd84dd3c550c549c8a762a7f54b21d/href">https://medium.com/media/6abd84dd3c550c549c8a762a7f54b21d/href</a></iframe><h3><strong>Portfolio</strong></h3><p>To make it simple, I’ve broken out my advice here into four themes:</p><h4>1. Make it easy to access</h4><p>This may sound silly, but the single most important thing about your portfolio is that I can open it. I would guess that 1/20 applicants to our jobs either neglect to include the link to their portfolio altogether or don’t share the password to it. When this happens, I’ll make a call, based on your resume, whether it’s worth the hassle to reach out to you for the link or password or not. Please just make it as easy as possible for me to look at your work. I know you may be trying to protect confidential client stuff but don’t do it at the expense of it being easy for me to access your portfolio.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FfhEJ9H5BcuRrr8eIgu%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2FHBOMax-hbomax-gossip-girl-on-2021-fhEJ9H5BcuRrr8eIgu&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2FfhEJ9H5BcuRrr8eIgu%2Fgiphy.gif%3Fcid%3D790b76117064c87f7c8a2aa8f7fd9c91439d20708cd68078%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/c707502475fd143010c989a5884e21c1/href">https://medium.com/media/c707502475fd143010c989a5884e21c1/href</a></iframe><h4><strong>2. Stand out from the crowd</strong></h4><p>Show off your design skills with the design of your portfolio site itself. This may seem obvious, but most candidates do not treat their portfolio site as the first product design of theirs I’m evaluating. Of course it should be easy to navigate but most importantly it should digitally embody you. It should give me a feeling of what it would be like to work with you. It’s great that it’s so easy to make a portfolio website these days, but as is true with web design overall, things have become homogenized. Does your portfolio site look identical to hundreds of others I’ve looked at or is it unique in some special way that helps me get a sense for who you are or what you care about? Show me who you are, not just that you can fill out a Squarespace template. Get weird with it! There is no requirement stating you must have a huge “Hi, I’m so and so and I’m a product designer.” intro on the landing page. In fact, maybe rethink the value of that bold landing page altogether! Your case study content is way more important than the landing page itself so why not guide me directly to some of your favorite work?</p><p>Another way you can stand out is to wow me with your beautifully crafted, exquisitely mocked-up final designs. You’d be amazed how many candidates seem strong all the way from first impressions through the entire process only to land at a pixelated, antiquated, downright poor quality final design mocked up in a blurry iPhone 5.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FnITKVYrWIatviLkfWN%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2FRoadshowPBS-nITKVYrWIatviLkfWN&amp;image=https%3A%2F%2Fmedia3.giphy.com%2Fmedia%2FnITKVYrWIatviLkfWN%2Fgiphy.gif%3Fcid%3D790b7611ff967a2ce153da2b33e08c9ced32fac965fa0f6f%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="271" frameborder="0" scrolling="no"><a href="https://medium.com/media/798561b6058f4df52c9fe1088d020810/href">https://medium.com/media/798561b6058f4df52c9fe1088d020810/href</a></iframe><h4><strong>3. Make the process your own</strong></h4><p>Show me <em>your</em> tailored design process, not just the formulaic process you were taught from a textbook or bootcamp. Help me understand how each step of your process helped you make decisions and informed your rationale. Product design is not a recipe you can just follow to get a delicious meal. We’re making dishes that have never been cooked or tasted before! To do that well, you need to be able to flourish in ambiguity, tap into your raw creativity, and adapt to constantly changing requirements.</p><p>At Cohere, nobody will tell you what process you have to follow, and it may not always be the same from project to project. The process of getting from initial problem discovery through to final designs and measurable impact can require you to use different tools. We need to know you can think critically about the design process and be thoughtful about which activities you do and why.</p><p>A common way I see candidates failing this part of our evaluation is in their presentation of personas that have no clear connection to the rest of their process or to the final design solution. While they can be an instrumental tool in better understanding the clusters of user behaviors and needs, I often see poorly thought-through personas which the candidate breezes over as if they are literally checking that step off the list.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FP2Y5CA2lJRmDbUux8T%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Ftherokuchannel-demi-lovato-roku-the-show-P2Y5CA2lJRmDbUux8T&amp;image=https%3A%2F%2Fmedia3.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExenA2dGFsODk0dzA2bnh5bWNmYTI1cHR3dGxla3owdHZ4Y25sMWc3dCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FP2Y5CA2lJRmDbUux8T%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="244" frameborder="0" scrolling="no"><a href="https://medium.com/media/a523e6d2e2dfd86cbd77c014c3848c1f/href">https://medium.com/media/a523e6d2e2dfd86cbd77c014c3848c1f/href</a></iframe><h4><strong>4. Explore wild ideas and rationalize your decisions</strong></h4><p>Show me a wide and creative array of solution ideas for any given problem. It could be high-level concepts or micro UI details. It doesn’t matter, as long as I can see that you thought about the problem and explored through sketching, wireframing, or pixel-pushing, many different ideas. Some may be obvious, some may seem insane! In fact, I love seeing the insane ideas! Aim to shock me with an idea you explored.</p><p>At Cohere, our first solution idea is rarely the one we end up building. Sometimes it takes exploring dozens of potential ideas before arriving at one that seems to sufficiently solve the problem. Our best design reviews are when our designers bring a set of concepts to a cross-functional team of stakeholders (PMs, engineers, clinical content writers, doctors, etc. — you get the idea), and together as a team, everyone evaluates what’s working and not working from each concept. Show me you can do that and you will stand out significantly from your fellow applicants.</p><p>It doesn’t matter if you’re showing me healthcare B2B workflows or an app to order pizza, I need to see the full array of ideas you explored to solve the problem and understand what it was about each idea you were interested in. Were you exploring innovative interaction models or just different placements of the primary button? Show it all! I am dismayed at how rare this is to find in portfolios.</p><p>Once you’ve shown off a wide variety of super creative ideas, help me understand how you made the decisions you did. Provide your rationale for why one idea was better than another. Yes, “we learned from user testing that they preferred option 1” or, “My PM said option 2 would better address the business need” are acceptable answers here, but they won’t help you stand out. We’re looking for candidates who have a strong point of view, who deeply believe there is something interesting about every one of their designs, and who can clearly articulate that rationale.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FxT0xem7ZlZ2DOYqpG0%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Flatenightseth-seth-meyers-lnsm-xT0xem7ZlZ2DOYqpG0&amp;image=https%3A%2F%2Fmedia1.giphy.com%2Fmedia%2FxT0xem7ZlZ2DOYqpG0%2Fgiphy.gif%3Fcid%3D790b76110ce67af8ebcbe592958b506ae49508e6b5c84a19%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/62c6c10bb9b7c4a6dff4a59a80088e25/href">https://medium.com/media/62c6c10bb9b7c4a6dff4a59a80088e25/href</a></iframe><h3><strong>First interview</strong></h3><p>Ok, so let’s say you made it past the initial screening with our amazing recruiter and you’re preparing for a 1:1 interview with the hiring manager, here are some key things to keep in mind:</p><ul><li>Unless your portfolio site is designed to support presentation-style storytelling, which it most likely is not, it will be much easier for you and your interviewer, if you prepare a slide deck.</li><li>Be prepared to walk through one, possibly two case studies, and be sure to excel at all 4 of the portfolio themes above as you present your work.</li><li>Be ready to ask questions that will actually help you decide if this is the right job for you. I’m amazed at how many candidates seem surprised that I give them time to ask me questions and as a result, have not prepared any. If the interview ends early because you can’t think of any questions to ask me, that is a huge red flag. I want to see intense curiosity about what it’s like to work here. I expect you to interview me in whatever ways are meaningful to you.</li></ul><h3><strong>Design exercise</strong></h3><p>I can’t speak for other company’s design exercises, but if you’ve made it this far and are about to start ours, here are some tips:</p><ul><li>We suggest you spend no more than 4 hours and we mean it! When I did the exercise (yes, everybody does it!), I spent the first 10 minutes planning out how I wanted to use the time, then I set a timer for myself so I’d know when I needed to switch to the next step of my process. It’s totally up to you how you want to use your time — I suggest you flex whatever part of the design process you’re most excited about.</li><li>Our exercise is like an entire end-to-end product design project shrunken down into a tiny fraction of the real deal. We know that can be stressful, but what we’re evaluating is your ability to understand a complex problem and generate creative solutions to the problem, independently, and under extremely tight time constraints. We suggest that you prepare a presentation (walking through a live Figma file is totally fine too!) that includes hand sketches showing a variety of ideas, low-fidelity wireframes walking through the prioritized workflow, and then a couple high fidelity mockups and/or prototype showcasing your visual design skills. We want to see you attempt to solve some of the complex interactions and also bring some visual sizzle with the UI.</li><li>We have the highest expectations of craft and intuition for our product designers, so it’s essential we get a firm grasp on how you think and work through problems on your way to arriving at elegant, beautiful solutions. We want to see the messy, scrambly, flailing process just as much as we want to see the polished final UI.</li><li>As you’re doing it and once you’re finished, think about whether you enjoyed the exercise or not. The idea is to see how you take prioritized user stories and design a product based on them, because that’s how we work here. If you hated it, this role probably isn’t a great fit for you. If you had a blast and just can’t wait to do more, you’re gonna love working on our team!</li><li>At this stage, we ask candidates to present their design exercise work to our design team as well as a representative from engineering and product management which is yet another opportunity to assess whether this team of folks feels like a good fit for you. The audience you’re presenting to is representative of who attends design reviews so this is a great ay to get a feel for how it will feel to present work to them after you join. Be prepared to answer questions from cross-functional collaborators because those can be different from what designers typically ask.</li><li>And once again, please come prepared with questions for the entire team. You should walk away knowing how the team functions and what you might be working on initially.</li></ul><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FS9ztGXw9d3o0SO738Q%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2FHBOMax-good-job-legendary-hbo-max-S9ztGXw9d3o0SO738Q&amp;image=https%3A%2F%2Fmedia0.giphy.com%2Fmedia%2FS9ztGXw9d3o0SO738Q%2Fgiphy-downsized-large.gif%3Fcid%3D790b76113e1a2a578f90d2aa4af1b516c84b9411440f065c%26rid%3Dgiphy-downsized-large.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="435" frameborder="0" scrolling="no"><a href="https://medium.com/media/73a69da9aae3fb0ac3fc2d5aba79af39/href">https://medium.com/media/73a69da9aae3fb0ac3fc2d5aba79af39/href</a></iframe><h3><strong>Final round</strong></h3><p>If you’ve made it this far, odds are you and the team are feeling really really good about this. We treat our final round as one last opportunity for you to meet some Cohere folks you may not have had the chance to meet yet or for you to spend more time with someone on the design team if you’d like to ask more questions. The only advice I’ll give here is to come prepared with any further questions you have that will give you as much confidence as possible that this is going to be a good fit for you. By the end of this process, you should have a deep understanding of what we do, why we do it, and why you want to be a part of it.</p><p>Oh and be yourself! We are a welcoming, empathetic, vulnerable bunch of folks and we can’t wait to get to know you better. Good luck!</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FxT5LMwN3B6j3OQcmUU%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fgiphy.com%2Fgifs%2Fthehills-xT5LMwN3B6j3OQcmUU&amp;image=https%3A%2F%2Fmedia3.giphy.com%2Fmedia%2FxT5LMwN3B6j3OQcmUU%2Fgiphy.gif%3Fcid%3D790b7611fc8b0c9e900770044596e96140e26349cb41e76b%26rid%3Dgiphy.gif%26ct%3Dg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="248" frameborder="0" scrolling="no"><a href="https://medium.com/media/eade19fff9ea56d03d2a9fbbbc5eb3cd/href">https://medium.com/media/eade19fff9ea56d03d2a9fbbbc5eb3cd/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ff08583dcdb7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/design-bootcamp/how-to-land-a-product-design-job-ff08583dcdb7">How to land a product design job</a> was originally published in <a href="https://medium.com/design-bootcamp">Bootcamp</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Dust Cloud of Nonsense]]></title>
            <link>https://medium.com/@sethdanielsnyder/a-dust-cloud-of-nonsense-a3fcc83c8731?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3fcc83c8731</guid>
            <category><![CDATA[truth]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[fake-news]]></category>
            <category><![CDATA[journalism]]></category>
            <category><![CDATA[social-media]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Tue, 17 Mar 2020 01:28:48 GMT</pubDate>
            <atom:updated>2020-03-17T01:28:48.882Z</atom:updated>
            <content:encoded><![CDATA[<h4><strong>Living and designing in a post-truth era</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5URNIXlvOjGI8qjdDtj-Uw.jpeg" /><figcaption>Various images of fake CGI Illustration: Guardian Design/AFP/Washington Post via Getty Images</figcaption></figure><p>You’ve no doubt heard quite a bit about “fake news” by now. It probably conjures memories of Facebook’s role in the 2016 U.S. election, or if you dug deeper, you would have read about the hordes of Macedonian fake news producers <a href="https://www.cnn.com/videos/cnnmoney/2017/09/12/the-macedonia-story-full-clean-lon-orig.cnn-money">who are now gearing up for the 2020 election</a>. If you know your journalism history, the term ‘fake news’ might remind you of the dozens of instances throughout history in which the spread of misinformation led to violence and war and helped dictators rise to power. And if you know your emerging tech, you’ve probably heard about <a href="https://motherboard.vice.com/en_us/article/594qx5/there-is-no-tech-solution-to-deepfakes">DeepFakes</a> and <a href="http://digg.com/2018/deepfake-what-is-it">OpenFaceSwap</a> and other AI-enabled media manipulation technologies. Whatever your context, I’m sure you’ll agree it’s getting harder and harder for us, as a society, to agree on or even know what’s truthful. And this is a really big problem. Let me tell you why:</p><p>We are living in an era of what President Obama called a “dust cloud of nonsense,” fueled by an epidemic of false information infecting the internet, using social networks as its catalyst.</p><blockquote><strong>“As long as it’s on Facebook, and people can see it, as long as it’s on social media, people start believing it. And it creates this dust cloud of nonsense,”</strong></blockquote><blockquote><a href="https://obamawhitehouse.archives.gov/the-press-office/2016/11/07/remarks-president-hillary-america-rally-ann-arbor-michigan">U.S. President Barack Obama in the run up to the 2016 US election</a></blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FcQ54GDm1eL0%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcQ54GDm1eL0&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FcQ54GDm1eL0%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/002d8915c14edd7968069d71086104ba/href">https://medium.com/media/002d8915c14edd7968069d71086104ba/href</a></iframe><p>We are living in a post-truth society in which fictitious clickbait beats out the facts. Technologies that can be used to modify and distort real information are evolving faster than our ability to understand, control, or mitigate them. We have created an environment where misinformation thrives and true information can confuse and paralyze rather than inform and illuminate. Technology has enabled us to make so many advances, but in the case of journalism, we’re moving the wrong way. We have created a dangerous combination of powerful, open-source neural network research, and an internet economy built on platforms and user experiences engineered to encourage viral spread, all of which is compounded by a population of people who are more likely to click “share” than to take the time to verify the veracity of the media they’re consuming.</p><blockquote><strong>“Our platformed and algorithmically optimized world is vulnerable — to propaganda, to misinformation, to dark targeted advertising from foreign governments — so much so that it threatens to undermine a cornerstone of human discourse: the credibility of fact.”</strong></blockquote><blockquote>Aviv Ovadya, Chief Technologist for the University of Michigan’s Center for Social Media Responsibility</blockquote><h3><strong>How did we get here?</strong></h3><p>Design and technology are both partially responsible for the current state of affairs, but they can also be tools to help us find an ethical path for the creation and spread of information that guides us away from the ‘post-truth’ future towards which we are currently bounding. We (<a href="https://medium.com/u/1e72b0be262b">Eve Weinberg</a> and <a href="http://sethsnyder.net/">Seth Snyder</a>, both UX designers who crossed paths at <a href="https://www.frogdesign.com/">frog design</a> in San Francisco) set out to understand the broader technology landscape and the history of “fake news” that got us to where we are today so that we could propose actionable opportunities for designers to guide us out of this post-truth era. But before we jump to these opportunities, let’s understand how we arrived at this “dust cloud of nonsense” we call home in 2020. As we see it, the problem can be broken down into three themes:</p><ol><li><strong>The emotionally muddled fact</strong></li><li><strong>Financial and political incentives</strong></li><li><strong>Technology-first innovation</strong></li></ol><h4><strong>1. The emotionally muddled fact</strong></h4><p>Facts are objective. They are proven with evidence. They can be demonstrated repeatedly. The word ‘fact’ can be used synonymously with truth, as distinct from opinions, beliefs, or matters of taste. However, according to the American Heritage Dictionary (perhaps not coincidentally published by the Heritage Foundation, a right-wing think tank), the word ‘fact’ also has a long history, dating back to the 4th edition in 1729, of being used as an allegation of something that may or may not be a true fact, as in “the author’s facts are not trustworthy” or, as Kellyanne Conway famously dubbed them ‘<a href="https://youtu.be/VSrEEDQgFc8?t=109">Alternative Facts</a>’.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FVSrEEDQgFc8%3Fstart%3D96%26feature%3Doembed%26start%3D96&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DVSrEEDQgFc8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FVSrEEDQgFc8%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/e1ae42f2b567657f5c6303d6d0ec47ce/href">https://medium.com/media/e1ae42f2b567657f5c6303d6d0ec47ce/href</a></iframe><p>As soon as a fact becomes something that has the possibility of being untrue, all of a sudden, one’s opinions and beliefs serve as sufficient evidence. Add to this mix, vast online communities of like-minded believers or disbelievers in whichever fact is currently up for debate, and you end up with socially constructed, shared assumptions about reality that <em>feel</em> a lot like facts.</p><p>Each of us needs to “construct” our own version of reality, and, in doing so, each of us is influenced by our communities, within which we have shared experiences and beliefs. Many media experts blame what has become known as the ‘Filter Bubble,’ the idea that, supported by digital customizations and microtargeting, we only tend to encounter our pre-existing beliefs online. In other words, our digital experiences in which we encounter what we perceive as facts may just be an echo-chamber of falsehoods. But, according to <a href="https://www.ted.com/talks/zeynep_tufekci_we_re_building_a_dystopia_just_to_make_people_click_on_ads">Zeynep Tufekci</a>, a techno-sociologist writing in the <em>MIT Technology Review Vol 121</em>, we actually encounter more alternative views thanks to the internet, than previous generations would have been exposed to. “The problem,” she says, “is that when we encounter opposing views in the age and context of social media, it’s not like reading them in a newspaper while sitting alone. It’s like hearing them from the opposing team while sitting with our fellow fans in a football stadium. Online, we’re connected with our communities, and we seek approval from our like-minded peers. We bond with our team by yelling at the fans of the other one. In sociology terms, we strengthen our feeling of “in-group” belonging by increasing our distance from and tension with the “out-group” — us versus them. Our cognitive universe isn’t an echo chamber, but our social one is. This is why the various projects for fact-checking claims in the news, while valuable, don’t convince people. Belonging is stronger than facts.” Furthermore, the design of online experiences leaves people with little time or resources to vet sources or question their validity.</p><p>But the internet, and the more rapid democratized spread of information that came along with it, couldn’t have been all bad for truth, you might be saying to yourself. And you would be correct. Whereas in the pre-internet days, the social construction of facts took place behind closed doors by old-world information gatekeepers such as the traditional media, NGOs, government, and academic institutions, our facts are now being forged in the same internet factories where we connect with our friends and loved ones. And, at first, we all thought this was a good thing. From the earliest days of social media, the Arab Spring was a shining moment in which we saw the potential of an open internet to liberate people from old-world censorship. But, while the weakening of the old-world information gatekeepers liberated and empowered the underdogs to speak out about oppression and connect with others for support, it also, in other ways, deeply disempowered them. As Tufekci puts it, “dissidents can more easily circumvent censorship, but the public sphere they can now reach is often too noisy and confusing for them to have an impact. Those hoping to make positive social change have to convince people both that something in the world needs changing and there is a constructive, reasonable way to change it. Authoritarians and extremists, on the other hand, often merely have to muddy the waters and weaken trust in general so that everyone is too fractured and paralyzed to act.” In other words, the burden of proof is on the person creating the fact. If someone wants to change a fact into something else, something they know will stir an emotional reaction from a group of people, all they have to do is question it, without the requirement to prove anything. Take President Trump, for example: his vigorous undermining and diminishing of journalism as having any relationship with the truth has no doubt created fertile ground for misinformation. Trump’s emotional argument that we can’t trust news organizations, leaves truth-telling no place to stand. The result is a net loss in the investigation of and accountability for those who exercise power and for truth in general.</p><p>But there is hope! A burgeoning movement among academics and media advocates suggest we, as a society, find ways to leverage the basic, yet immensely powerful emotion, shame, to change behavior for good. Harvard researcher Claire Wardle admits that tech companies have a big role to play in stemming the tide of misinformation but that the problem cannot be solved by engineers alone. She and others are calling for an intervention that makes sharing misinformation as shameful as drunk driving or as embarrassing as littering in public. By helping people see the aggregate effect of their re-sharing of “fake news” posts on social media, for instance, they will understand their contribution to the web becoming a more toxic place. Proposals like this understand that we can fight emotionally charged language with emotions because you simply can’t reason against emotion. As much as it may seem that objective, trustworthy information sources are at odds with emotional hubs of human belonging, as designers, we must reconcile that tension and create information platforms and that deliver on both needs. The idea is, rather than focus strictly on technological solutions to combating “fake news,” we need to consider and design for the social forces that drive how we use technology. The goal being a citizenry that responsibly ingests information and can parse emotionally charged misinformation from objective facts.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F3iD4HwJ-67Q%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D3iD4HwJ-67Q&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F3iD4HwJ-67Q%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/7bbf8506b7ffc4a3aea1aace817857e9/href">https://medium.com/media/7bbf8506b7ffc4a3aea1aace817857e9/href</a></iframe><blockquote><strong>“Fake news drips drops of poison into our daily web diet and we end up infected without even realizing it. It’s only right to give these kids the possibility to defend themselves from lies.”</strong></blockquote><blockquote>Laura Boldrini, president of the Italian lower house of Parliament, who has spearheaded a project with the Italian Ministry of Education to train students how to recognize fake news and conspiracy theories online</blockquote><h4><strong>2. Financial and political incentives</strong></h4><p>Underlying all of these addictive online experiences and our desire to belong, are a powerful set of incentives to nudge large populations towards beliefs that will lead to a specific outcome, whether that’s buying something or voting for a particular candidate. And these lucrative incentives fuel an economy of nefarious propagandists who profit off the creation of made-up news stories, conspiracy theories, and social media accounts. But “fake news” is not a new phenomenon, nor are the financial and political incentives that fuel it. It has been present since news became a concept five hundred years ago with the invention of print — a lot longer, in fact, than verified, “objective” news, which emerged in force a little more than a century ago. From the start, “fake news” has tended to be sensationalist and extreme, designed to inflame passions and prejudices, all the while, leveraging its grip on our emotions for financial gain. And in many cases, it has had violent results, from inciting gun violence at a fake Hillary Clinton child abuse ring to the spark of the Spanish-American War, to the massacre of Jews in the 1400s, and Nazi propaganda that built on the same prejudicial lies.</p><blockquote><strong>“You furnish the pictures and I’ll furnish the war.”</strong></blockquote><blockquote>William Randolph Hearst</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5YdHrMgWL7ElERH4gkjczg.jpeg" /></figure><p>The new generation of algorithmic information gatekeepers, platform tech companies like Facebook, Twitter, and Google are not that different in the way they set the stage for the interrelated incentives of politics and profit. Unlike most modern news organizations that are motivated to provide truthful content, platform tech companies (that are staunchly against the idea that they are news organizations despite growing numbers of us who site social media as our primary news source), claim that they are not responsible for the information other people post on their platforms. And they are legally within their rights to do so, thanks to Section 230 of the Communications Decency Act (CDA) of 1996. But while this law has frequently been referred to as a key to the flourishing of the Internet, we now know that information on these platforms that is misleading and polarizing yields the greatest financial rewards. And political content has proven to be some of its most addictive and toxic misinformation. <a href="https://www.wsj.com/articles/how-youtube-drives-viewers-to-the-internets-darkest-corners-1518020478">A Wall Street Journal investigation</a> found that YouTube’s recommendation algorithm tended to drive viewers toward divisive, misleading or false content and what was often an edgier version of whatever they were watching — a good way to hold their attention. These platforms have expressed minimal interest in pumping the breaks or shifting their business models (with a few exceptions being <a href="https://www.facebook.com/facebookmedia/blog/working-to-stop-misinformation-and-false-news">Facebook’s internal efforts to curb fake news</a> or <a href="https://www.washingtonpost.com/technology/2018/07/06/twitter-is-sweeping-out-fake-accounts-like-never-before-putting-user-growth-risk/?noredirect=on&amp;utm_term=.9301edb95a4c">Twitter’s removal of 70 million bot accounts</a>), and our current government seems incapable of even fathoming a reasonable set of regulations. <a href="https://www.engadget.com/2018/04/11/zuckerberg-hearings-wasted-opportunity/">The senators in the Zuckerberg congressional hearing certainly didn’t make a dent</a>. Through the prioritization of clicks, shares, views, and microtargeting over factual content, the platforms have become some of the most profitable companies in the world, all the while shielded by Section 230.</p><p>But this isn’t exactly a new problem. Media historians know well that journalistic success has always been deeply entangled with financial and political incentives and has also had a complicated relationship with the truth. In the beginning, newspapers were almost entirely partisan, funded by wealthy individuals and political parties who used the medium to manipulate the public into electing their preferred candidates. Then came the Penny Press — cheap, independent newspapers who insisted they were operating in the public good, free from political ties, dedicated to giving people the facts on the ground. The role of “reporter” was born — ordinary working class people who went to crime scenes to investigate and wrote stories catering to what they thought people wanted to read. This established a chain of trust between local, state and national reporters and the public that had never existed before. That trust quickly began eroding, however, as newspapers began to treat their audiences like consumers — they imagined what their audiences might like to read (and purchase) and crafted poetically embellished stories with a blatant disregard for the truth. The competition between William Randolph Hearst and Joseph Pulitzer’s media companies in New York took what was already a wild west of journalism with very few rules, and cranked up the “fake news” volume with the Yellow Press, known for its sensational creativity. Faking, at this time, most often didn’t mean nefarious manipulation like it does today. Rather, it meant a conscientious embellishment aimed at giving readers a better experience, by immersing them in a more engaging truth, closer to what they expected. It was an intentional design decision, for an improved customer experience that was taught and advocated by the top journalists of the time. Eventually, though, people began to think of journalists as incapable of telling the truth. It reached a crisis point. It was at this low point, at the beginning of the 20th century, that a shift started happening, towards objectivity, neutrality, and accuracy, led by the New York Times. Of course, the Yellow Press continued to thrive as well, and still does today, in our grocery store checkout aisle tabloids and online.</p><p>The challenge from the very beginning has been to find a way to fund the not inexpensive process of uncovering facts, reporting on them, and distributing them to a wide audience in a way that does not incentivize individual reporters and information distributors to bend the truth. The incentive of a news organization should not be to make money; it should be to uncover and publish the facts. Yet, unsurprisingly, we see business tycoons like Rupert Murdoch and large conservative organizations like Sinclair Broadcast Group trying to both monetize and control the message for financial and political gain.</p><p>Advertising and its associated revenue have always played a role in the success of information and social platforms. But our modern digital tools allow for advertisements to be highly personalized and emotional, micro-targeted to say exactly what will resonate with each of us. Ads are intentionally designed to blend in with the content such that they can best masquerade as trustworthy information. The financial incentives are the same as they’ve always been, but now they are amplified and enhanced by technology and design. And those same financial incentives are also fed by people and organizations who are more than happy to pay to target individuals who may be easily manipulated towards a purchase or a vote.</p><p>The key to turning this around, we believe, lies in shifting the financial incentives that have shaped our new generation of algorithmic information gatekeepers into conduits for misleading and polarizing information. Whether that means new social networks that don’t rely on microtargeted advertising or new business models for investigative journalism that allows it to focus on truth-telling rather than selling. We, as a society, must find ways to leverage the same energy that started the backlash against Yellow Journalism and sent the public in search of more objective news.</p><h4><strong>3. Technology-first innovation</strong></h4><p>Emerging technologies are making it easier than ever to manipulate perception, falsify reality, and spread lies. The tools to do so are becoming easier to use, more widely available, and may soon integrate seamlessly into our daily social workflows. Worse still, the tech companies and research labs behind them have no apparent benevolent use-case in mind or <em>only</em> consider fictional benevolent use cases rather than spending the time on stress testing nefarious use cases and building safeguards around them upfront. Media experts are calling for a change. If we don’t do something soon, they say, the future could become an “Infocalyptic Dystopia,” signaled by an increasing erosion of trust in media of any kind. This future, as described in <a href="https://www.buzzfeednews.com/article/charliewarzel/the-terrifying-future-of-fake-news">BuzzFeed</a> by Aviv Ovadya, chief technologist for the University of Michigan’s Center for Social Media Responsibility, will be “Worse because of our ever-expanding computational prowess; worse because of ongoing advancements in artificial intelligence and machine learning that can blur the lines between fact and fiction; worse because those things could usher in a future where anyone could make it “appear as if anything has happened, regardless of whether or not it did.”</p><p>If you’ve been paying close attention to developments in AI and machine learning, this won’t feel like much of a stretch. From <a href="https://www.theverge.com/2019/3/19/18272602/ai-art-generation-gan-nvidia-doodle-landscapes">Nvidia software</a> that can generate hyperrealistic photos of objects, people, and even some landscapes, to the Adobe projects — Voco and <a href="https://research.adobe.com/news/cloak-remove-unwanted-objects-in-video/">Cloak</a> — the first a “Photoshop for audio,” the second a tool that can seamlessly remove objects (including people!) from video in a matter of clicks, to the real-time face tracking video manipulation being developed at Stanford and the University of Washington known as <a href="https://www.theguardian.com/technology/2020/jan/13/what-are-deepfakes-and-how-can-you-spot-them">DeepFakes</a>, we are in the midst of a growing tidal wave of media manipulation tech.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fp5U4NgVGAwg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dp5U4NgVGAwg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fp5U4NgVGAwg%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/356b2be30b91c69b00bb4babe24717c8/href">https://medium.com/media/356b2be30b91c69b00bb4babe24717c8/href</a></iframe><blockquote><strong>“As computer images get better, we believe all images less.”</strong></blockquote><blockquote><a href="https://vimeo.com/237568588">https://vimeo.com/237568588</a></blockquote><p>There is an irony, points out Shankar Vedantam, <a href="https://www.npr.org/2018/06/25/623231337/fake-news-an-origin-story">in this Hidden Brain podcast episode</a>, in the fact that, over and over throughout history, “people tend to assume new technologies will make it harder to manipulate the truth. But in reality, new technologies — from photography to the Internet — open up new avenues for faking.” History has shown us that new technologies make us more susceptible to manipulation, yet we fall for it time after time. Let’s look at a hallmark example:</p><p>When the daguerreotype, the first publicly announced and commercially viable photographic process, was introduced to the world in 1839, it took the world by storm. Early photography was seen as the most realistic depiction of the truth ever made possible. It was universally known to be true that if it was captured in a photograph, it must have happened exactly as pictured. But just as soon as this social construct of truth reached its peak, an enterprising Boston man named William H. Mumler invented a process of doubling the exposures on a single negative to create ghostly blurs of deceased people, that became known as Spirit Photography, capitalizing on the public belief that this new magical technology could do anything. With the support of his studio’s receptionist, who freelanced as a medium, Mumler immediately captured the attention of a national audience and he received visitors from across America, including the recently widowed Mary Todd Lincoln. At a violent time in America when people were desperate to connect with their deceased loved ones, Spirit Photography, despite its overt trickery and fraud, was a tangible symbol of the overarching argument of mysticism and fake-ness as a desired product. Its ability to resonate with people emotionally is what made it so successful, versus science and rationalism that did little to comfort the grieving.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/968/1*W8ouQJ7zoSY7rh2Fx90sWA.png" /><figcaption><em>Mary Todd Lincoln photographed with President Lincoln’s ‘spirit by </em>William Mumler<em> in 1870 — </em><a href="https://www.history.com/news/spirit-photography-civil-war-william-mumler">https://www.history.com/news/spirit-photography-civil-war-william-mumler</a></figcaption></figure><p>In the chilling RadioLab podcast episode in which Jad Abumrad and Robert Krulwich reported on the work at Adobe and elsewhere to manipulate video and audio in ways that, to an outsider, must appear equally as magical as Mumler’s Spirit Photography did in its time, they ask the all-important question: <strong>why?</strong> Video manipulation tech that lets anybody easily put somebody else’s face into a video; <a href="https://techcrunch.com/2017/04/25/lyrebird-is-a-voice-mimic-for-the-fake-news-era/">Lyrebird’s</a> Photoshop for audio that lets you recreate anyone’s voice and make them say things they didn’t. Why build these tools that have such clear potential to cause harm? In fact, we’ve already seen plenty of damage done by the nefarious use of tools like these.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FttGUiwfTYvg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DttGUiwfTYvg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FttGUiwfTYvg%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/32dceb405e40f2b9e21b285d3559c0d3/href">https://medium.com/media/32dceb405e40f2b9e21b285d3559c0d3/href</a></iframe><p>Not surprisingly, we’re seeing this tech used in the <a href="https://motherboard.vice.com/en_us/article/bjye8a/reddit-fake-porn-app-daisy-ridley">porn industry</a>, to create videos that realistically superimpose the faces of celebrities on the adult actors’ bodies. And as machine learning technologies advance, even more complex deception will become possible, and fakery will become even harder to spot. It may soon become so good that it will be impossible to tell the difference between an AI-generated face swap and the real thing. According to Will Knight, MIT Technology Review’s Senior Editor for Artificial Intelligence, “these advances threaten to further blur the line between truth and fiction, adding to the pervasive dissemination of disinformation through fake social media accounts accelerated by the internet. “Alternative facts” and conspiracy theories are already widely believed. Now imagine throwing real-looking fake videos into the mix: politicians mouthing nonsense or ethnic insults, or getting caught behaving inappropriately on video — except it never really happened.” In an interview with Charles Seife, a professor at NYU and the author of Virtual Unreality: Just Because the Internet Told You, How Do You Know It’s True? Knight reports that Seife confessed to being astonished at the pace of progress since his book was published in 2014. “Technology is altering our perception of reality at an alarming rate,” he said.</p><blockquote><strong>“If you can’t tell a fake from reality, then it becomes easy to question the authenticity of anything.”</strong></blockquote><blockquote>Will Knight, MIT Technology Review’s Senior Editor for Artificial Intelligence</blockquote><p>What’s more, the mere existence of this technology is enough to cause harm, even if it’s rarely used. As we saw with <a href="https://www.nytimes.com/2017/11/28/us/politics/trump-access-hollywood-tape.html">Donald Trump’s denial that it was his voice</a> on the infamous <em>Access Hollywood</em> tape, citing experts who told him it’s possible it was digitally faked, the possibility itself was enough to denounce the credibility of the media that should have implicated him. As computational propaganda researcher Renee DiResta puts it, “you don’t need to create the fake video for this tech to have a serious impact. You just point to the fact that the tech exists and you can impugn the integrity of the stuff that’s real.”</p><p>Stanford University, The University of Washington, Adobe; these are clearly not organizations known for purposely creating tools for those with malicious intent. Yet somehow, researchers and software developers at these places and many others all over the world, independently reached the conclusion that it was a good idea to make these machine learning tools more widely available, making it possible for anyone with a decent computer and a few hours to spare to create a convincing face-swap that would have previously required a professional movie studio with millions of dollars. Why? Is our societal ambition to test the limits of what’s possible with technology so grand that we sometimes ignore the risks of such technological boundaries? Sure we can more easily re-shoot footage for movies or edit audio for podcasts or make cool music videos like <a href="https://www.theverge.com/2018/10/11/17963166/charli-xcx-1999-video-spice-girls-deepfakes-ryan-staake">this one</a> or, as one of the graduate students at the University of Washington who created <a href="http://grail.cs.washington.edu/projects/AudioToObama/siggraph17_obama.pdf">the lip sync technology</a> says in <a href="https://www.ted.com/talks/supasorn_suwajanakorn_fake_videos_of_real_people_and_how_to_spot_them?language=en#t-411338">his TED talk</a>, create virtual simulations of ourselves or of dead loved ones to “allow any individual’s positive impact on the world to be massively scaled.” But at what cost?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F6-v1b9waHWY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D6-v1b9waHWY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F6-v1b9waHWY%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/c8b22bdd712a8b518fa08724d670f967/href">https://medium.com/media/c8b22bdd712a8b518fa08724d670f967/href</a></iframe><p>Hearing about these non-human-centered, solution-first approaches to technology development reminds us of the importance of starting the process with a true human need, and maintaining a focus on that need throughout the development of a new technology. As members of the creative community, we must keep each other accountable to a standard of ethics that helps us weigh the value of solving a human need with the potential negative impacts of the solution. In this case, the time savings for low-budget movie producers who can now simulate their actor’s voice if they didn’t get quite the right take, against the global rampage on truth caused by this technology becoming mainstream. When technology is developed in a vacuum, without a true human need to address, or without proper consideration of how the tool might be used maliciously, it can easily be abused or used to cause harm. And when those technologies become so easy to use and disseminate, it becomes more important than ever to remain human-focused and ask, ‘What is good for society? What is worth making that will help, and not hurt humanity?’</p><p>In the meantime, there may be some hope in the leveraging of the very same machine learning mechanisms used to conjure these ever-more convincing deceptions to also detect them in the wild. Building upon decades of well-established methods for identifying doctored images and videos, media cryptographers are developing advanced techniques to detect visual inconsistencies and physical oddities that signal a manipulated piece of media. And since most people can’t do this kind of detective work on their own and don’t have enough time to study every image or video clip that pops up on their Facebook feed, there has been a push to automate the detection of fakery. There is also an ever-growing need to double down on our collective efforts to establish safeguards that protect the rights of citizens who fall victim to nonconsensual simulacra. Lyrebird claims that in the future it will limit its voice simulations to people who have given their permission, but can we count on everyone being so scrupulous? In the end, tools are simply tools and anything can be used for good or evil. But as tool makers, designers can help others imagine nefarious use cases and help design better tools (link to Superflux). And as experience creators, we can try to shift behaviors and social norms towards benevolent use cases. These are by no means easy tasks, but the future of truth depends on it.</p><h3>Breaking through the dust clouds</h3><p>So yeah, in a lot of ways, we’re already living in a Post Truth society. But that’s certainly no reason to give up! These are mere constraints. To a designer, problems are opportunities and constraints are solicitations for creativity. It’s better design vs. total information dystopia, people! Better design of the information delivery mechanisms we all know and love, better design of our academic research centers where emerging tech is so often born, better design of information itself. We must intervene before we let society descend into total reality apathy, in which the torrent of constant misinformation makes us all simply start to give up. Which by the way, is a well-documented tactic of authoritarian leaders.</p><p>Now we’d like to explain how we as designers can take action. As human-centered user-experience designers, we approach problems by first finding the opportunities for change. What we intend is for the opportunity areas below to guide fellow designers through generating concepts for products, services, policies, etc. If you’re new to the ideation process, we recommend the following structure:</p><ol><li>Gather 6–10 people who you think would bring an open-minded, creative perspective to the problem into a room for at least an hour. Assign a facilitator to guide the brainstorm, take notes, and keep time.</li><li>Select one of the ‘How Might We’ statements from the lists below and begin generating ideas for tangible, implementable solutions. As soon as the group begins to slow down, move on to another HMW statement, and continue in this pattern for as long as you like.</li><li>At the end, review the ideas you generated and prioritize a few that feel like they would have the biggest impact and are easiest to implement.</li><li>Start building them!</li></ol><p>Ready? Here goes! Let’s think about these opportunities to combat fake-news in three parts along a journey:</p><ol><li><strong>CREATION — the birth of new content</strong></li><li><strong>CLASSIFICATION — determining if the content is factual or fictional and communicating the result</strong></li><li><strong>DISSEMINATION — spreading the content into the world</strong></li></ol><p>Imagine, if you will, a piece of news traveling along this journey (not unlike everyone’s favorite Schoolhouse Rock song about the <a href="https://www.youtube.com/watch?v=FFroMQlKiag">Bill on Capitol Hill</a>) and our goal as designers is to help our piece of news live a better life and make better choices. At each stage of the journey, some promising solutions have emerged over the past few years, and we’ll do our best to include them here, for reference. In many cases, these fledgling projects need great designers and thinkers to take the idea to the next level, so we consider those just as fruitful as net new ideas.</p><ol><li><strong>CREATION — the birth of new content</strong></li></ol><p>The opportunity areas we list here suggest a simple ‘carrot and stick’ motivational framework. Some are meant to encourage positive outcomes (carrot) and some are meant to punish negative outcomes (stick). We encourage ideation of both.</p><p><em>Opportunity Areas</em></p><ul><li>How might we (HMW) establish new business models for investigative journalism so that they can continue holding the powerful accountable to the truth?</li><li>HMW encourage journalists and platform users to create content with a moral code or other incentive system that weighs truth above all else?</li><li>HMW disincentivize malicious content creators from making untruthful content in the first place?</li><li>HMW help content creators simulate the potential ramifications of their work?</li><li>HMW ensure researchers and technologists adhere to a code of ethics, with those ethics guiding future creations?</li><li>HMW create tech platforms and social networks that foster alignment and community, rather than fomenting polarization?</li></ul><p><em>Reference Projects</em></p><ul><li><a href="https://accountablejournalism.org/about">A collaborative project that aims to be the world’s largest collection of ethical codes of conduct and press organizations</a></li><li><a href="http://humanetech.com/">Design ethicist Tristan Harris’ Center for Humane Technology</a></li><li><a href="https://frontpageafricaonline.com/">FrontPage Africa, a crowd-funded independent newspaper in Liberia</a></li><li><a href="https://www.propublica.org/about/">ProPublica</a></li></ul><p><strong>CLASSIFICATION — determining if the content is factual or fictional and communicating the result</strong></p><p>These opportunity areas are the most actionable. They are ripe for talented designers to create tangible design solutions and products.</p><p><em>Opportunity Areas</em></p><ul><li>How might we (HMW) contribute to the determination of the truthfulness of content?</li><li>HMW we design tools to support automated media cryptography to verify the authenticity of images, video, and audio?</li><li>HMW teach consumers to effectively recognize fake news?</li><li>HMW most effectively visualize and communicate the results of this classification (or the likelihood that the content is truthful)?</li><li>HMW help truly unbiased fact-checkers communicate their own trustworthiness to skeptical outsiders?</li></ul><p><em>Reference Projects</em></p><ul><li><a href="https://towardsdatascience.com/i-trained-fake-news-detection-ai-with-95-accuracy-and-almost-went-crazy-d10589aa57c">Home-made machine learning model detects fake news with 95% accuracy</a></li><li><a href="http://nymag.com/intelligencer/2016/11/heres-a-browser-extension-that-will-flag-fake-news-sites.html">Chrome Extension That Will Flag Fake-News Sites for You</a></li><li><a href="https://www.forbes.com/sites/emilywillingham/2016/11/28/a-scientific-approach-to-distinguishing-real-from-fake-news/#4d2640442bd8">A Scientific Approach To Distinguishing Real From Fake News</a></li><li><a href="https://www.nytimes.com/2017/10/18/world/europe/italy-fake-news.html">An Italian education initiative to teach school students how to detect fake news online</a></li><li><a href="https://newslit.org/get-smart/">The News Literacy Project</a></li><li><a href="https://www.cs.tau.ac.il/~tromer/papers/photoproof-oakland16.pdf">Cryptographic Image Authentication Research</a></li><li><a href="https://ifcncodeofprinciples.poynter.org/">The International Fact-Checking Network</a></li><li><a href="https://www.buzzfeednews.com/article/craigsilverman/facebook-and-fact-checkers-fight-fake-news">Facebook’s internal initiative to partner with fact checkers</a></li><li><a href="https://www.factcheck.org/">FactCheck.org</a></li><li><a href="https://www.politifact.com/subjects/fake-news/">PolitiFact’s Fake News Label</a></li><li><a href="https://www.snopes.com/">Snopes</a></li></ul><p><strong>DISSEMINATION — spreading the content into the world</strong></p><p>These opportunity areas appeal to our society’s herd mentality. They are based on the well-known premise that social pressure can change behavior and shame can be a powerful force for good.</p><p><em>Opportunity Areas</em></p><ul><li>How might we (HMW) incentivize the spread of truth to be more salacious and spreadable than lies?</li><li>HMW help scientists, investigative journalists, and other creators of truthful content make their content more tangible, accessible, engaging, and shareable?</li><li>HMW discourage people from sharing untruthful content (public shaming)?</li><li>HMW make it harder to share content with a lower truthfulness rating?</li><li>HMW encourage the world’s largest information gatekeepers to prevent the spread of personalized disinformation?</li></ul><p><em>Reference Projects</em></p><ul><li><a href="https://www.facebook.com/help/1952307158131536?helpref=faq_content&amp;ref=pod">Facebook moves false stories lower in News Feed and notifies you if you share them</a></li><li><a href="https://ia.net/topics/domo-arigato-mr-roboto-tell-us-your-secret">Concept project to discourage the spreading of bot-generated content online</a></li><li><a href="http://time.com/5362183/the-real-fake-news-crisis/">Academic research into leveraging shame to prevent the spread of fake news</a></li></ul><h3>Carry on, truth warriors</h3><p>With an understanding of how we came to be living in a post-truth era, we hope that some of these opportunity areas inspire you to build solutions that get us one step closer to a truth-centric society. As with any pervasive problem of this scale, the way forward is to use our powers as designers and citizens to reshape how our tech companies, academic institutions, news organizations, media, and governments should function in the 21st century.</p><p><em>This essay has been over a year in the making, and we would like to thank our heroic editors Sherryl Goodman, </em><a href="https://medium.com/u/d378620715fd"><em>Agnes Pyrchla</em></a><em>, and </em><a href="https://medium.com/u/1b5c703dcf07"><em>Daniel Ishofsky</em></a><em>.</em></p><p><em>Please leave ideas and feedback in the comments!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3fcc83c8731" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Artificial Love: Design for Emotional Intelligence]]></title>
            <link>https://becominghuman.ai/artificial-love-design-for-emotional-intelligence-a5dd41a5202f?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/a5dd41a5202f</guid>
            <category><![CDATA[design]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[industrial-design]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Sat, 21 Jul 2018 02:03:51 GMT</pubDate>
            <atom:updated>2018-07-24T13:59:50.265Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7ijU09GStexTeIAeUZjXjw.gif" /><figcaption>Pearl Studios: <a href="https://pearlstudios.com/articles/robots-and-ai-need-love-too">https://pearlstudios.com/articles/robots-and-ai-need-love-too</a></figcaption></figure><h4>Part Three: <strong>Embodying Personality</strong></h4><p><strong><em>Note:</em></strong><em> This is Part 3 of a multi-part series looking at design opportunities for Emotional AI. Part 1 is </em><a href="https://becominghuman.ai/artificial-love-designing-for-emotional-intelligence-5b21c11242c8"><strong><em>here</em></strong></a><em> and Part 2 is </em><a href="https://becominghuman.ai/artificial-love-design-for-emotional-intelligence-9fa2966eb73b"><strong><em>here</em></strong></a><em>.</em></p><p>Welcome to Part 3, the final chapter of the Artificial Love series. If you’re reading each part back to back and you’ve just come from Part 2, feel free to skip down to the Embodying Emotional Connection section below. In Part 1, we discussed the growing shift in the way we interact with technology — from commands to conversations and how that shift will reshape our daily experiences with our products. We argued that designers have the potential to create AI products that we relate to in ways that resemble our favorite relationships with friends, relatives, co-workers, and pets. We presented frameworks to help you design products with such relationships in mind, with specific implications for how designers craft the ways that AI products behave, how much control designers have over our products’ personality and behavior. Finally, we proposed a set of aspirational models for the personality of a brand or an AI. In Part 2, we built on the premise that our evolving emotional connections with others are contingent on increasingly better understanding the other’s needs. We investigated how those connections may look when one member of the relationship is an emotionally intelligent AI product, such as a chatbot, robot, or autonomous car. And we provided guidance on how to design products with artificial emotional intelligence that get know us better over time, predict our needs, understand our emotional states, help to alleviate our anxiety, commiserate with us over bad news, and enhance our excitement and joy. In this Part 3, the final installment in this series, we will look at the critical decisions to be made around embodiment — the size, shape, and other physical characteristics of your AI product and how those decisions bring the promise of artificial emotional intelligence to life. We’ll also discuss the opportunities to explore with embodiment mechanisms such as the AI living solely inside one object, versus in the cloud, accessible from anywhere. Finally, we will address the elephant in the room — gender.</p><figure><a href="https://www.eventbrite.com/e/chatbot-conference-tickets-45884424549?utm-medium=discovery&amp;utm-campaign=social&amp;utm-content=attendeeshare&amp;aff=escb&amp;utm-source=cp&amp;utm-term=listing"><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oyftTzPL9qaCYc8z2EbcHg.png" /></a></figure><p>Let’s begin with a brief summary of some of the key points from the first two parts. It may be hard to imagine it today, but one day in the not-too-distant future, super-sophisticated intelligent AI beings will develop complex bonds with humans. What doesn’t take imagination is the fact that already today, we are seeing people form unidirectional relationships with their virtual assistants like Siri, and with robots like the Roomba (not to mention sex robots like <a href="https://realbotix.com/">Harmony</a>). Despite the fact that the feelings aren’t yet able to be reciprocated, we can already see the power these relationships can have over the people in whose pockets, desktops, homes, and cars they live. And this is just the beginning. The technology available today is a mere beacon of things to come — akin to primordial humans. As the technology advances from the single task Artificial Narrow Intelligence of today towards Artificial General Intelligence (AIs that are as smart as humans), our products will be able to read our facial expressions, answer abstract questions, creatively express themselves, and many other as yet undetermined abilities. We will finally feel as though our products actually understand us, as humans understand each other, unlocking not just practical abilities like anticipating our needs, but also the emotional support that comes with meaningful human relationships. In the meantime, it is critical that designers working in this space understand how their decisions impact the relationship that users form with their products.</p><h3>Top 3 Most Popular Ai Articles:</h3><blockquote><a href="https://becominghuman.ai/from-perceptron-to-deep-neural-nets-504b8ff616e">1. From Perceptron to Deep Neural Nets</a></blockquote><blockquote><a href="https://becominghuman.ai/neural-networks-for-solving-differential-equations-fa230ac5e04c">2. Neural networks for solving differential equations</a></blockquote><blockquote><a href="https://becominghuman.ai/turn-your-raspberry-pi-into-homemade-google-home-9e29ad220075">3. Turn your Raspberry Pi into homemade Google Home</a></blockquote><h3><strong>Embodying Emotional Connection</strong></h3><p>Designers and product developers know well that the physical form of the objects we connect with emotionally is a critical aspect of why we love them. The physical form of an object gives us a point of reference, something to look at, to point at, something tangible to blame when things don’t work out or to reward with love when they do. Therefore, we must ask what physical form your AI should take. To design the form of something is to listen to the object’s purpose, and to create something that feels natural for what it wants to be. It must satisfy the dual objectives of ergonomics and visual appeal. Here at <a href="https://www.frogdesign.com/">frog design</a>, our founder Hartmut Esslinger famously updated the classic Industrial Design adage “Form follows function” which represented the process of understanding the guts of an object and how it works, so as to design a better version of the object leveraging those constraints. Our revision, “Form follows emotion” builds on its predecessor to highlight our focus on creating products and experiences that people love, not just at first glance, but over time as the relationship with the product evolves.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplayer.vimeo.com%2Fvideo%2F59679411%3Fapp_id%3D122963&amp;dntp=1&amp;display_name=Vimeo&amp;url=https%3A%2F%2Fvimeo.com%2F59679411&amp;image=https%3A%2F%2Fi.vimeocdn.com%2Fvideo%2F446895399_1280.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=vimeo" width="1280" height="720" frameborder="0" scrolling="no"><a href="https://medium.com/media/2d5a95accfa38201068ab125d736b485/href">https://medium.com/media/2d5a95accfa38201068ab125d736b485/href</a></iframe><p>Now that a foundational understanding of how function can influence form is table stakes for being an industrial designer, what separates good industrial design from exceptional? We believe it starts with a deep understanding of not only the functional needs of our users but also their emotional needs. Of course, the thing must work well and be visually appealing, but it must also connect with users emotionally. Through answering questions like “What is the relationship between the user and the object?” and “How does the object fulfill unmet and unspoken emotional expectations?” we frogs recognize the value and potential of considering emotion as a driver in the design process. As our products develop the ability to understand and express realistic emotion, the forms they take will necessarily be driven by the emotional needs of the humans who choose to adopt them. A robot designed to care for the elderly, therefore should have a dramatically different form than a vacuum cleaning robot for young professionals, not only because they have different sensors and motors inside, but because the emotional needs of an elderly person and a young professional are unique. As you are developing ideas for the physical shape of your embodied AI, it is crucial to understand the latent emotional needs of your potential customers.</p><p>But physical shape is truly only one way to think about embodiment. Any industrial designer will tell you there is a rigorous process involved in designing a physical thing that begins with roughly laying out the components, evolves into form explorations sometimes referred to as gestures, and finally, is refined with considerations of color, material, and finish, known collectively as CMF. Although there are certainly no shortcuts through this process, and a professional industrial designer should be brought on to provide these services, there are some ways you can set that designer up for success before they even get started. The first is to consider balance. When exploring form directions, it is best to seek a harmonious balance between the form and the color, material, and finish. For instance, if the form is highly complex or detailed, the CMF should be simple or neutral to provide balance. Similarly, if the form is neutral or basic, a more exciting CMF will help it come alive. The same principle applies with the technology embedded in the product — the more complex or futuristic the technology, typically the simpler the form and CMF ought to be.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KrF07lhEQFdVda3LkzqeSQ.png" /></figure><p>With the popular AI embodied products on the market today (Amazon Echo, Google Home, Apple HomePod, Roomba, etc.) there is a clear pattern: since the technology is highly advanced, difficult for the average consumer to grasp, and even potentially frightening, the embodiments are all solidly on the simple side of the scale, to provide balance. They all share basic geometric forms like tubes, spheres, pucks, and blobs, that are minimally finished with solid colors, simple materials, and subtle finishes. Blobby, geometric consumer robots like Kuri, EllieQ, and Zenbo also seem to be following a similar trend, although there are some anthropomorphic exceptions. To oversimplify the status quo, AI technology and robots in the home are new phenomena. Newness can be perceived as alien. Familiar forms and materials wrap the potentially frightening technology in a comfortable shell that plays nicely with archetypes that can be found in the home.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qTHHBVbgTKOKXcw1VSIoCA.png" /><figcaption>Kuri, EllieQ, and Zenbo</figcaption></figure><p>As you develop the form of your new AI embodied product, ask yourself and your team:</p><p>· Where do you see your product on the scale of Traditional to Futuristic?</p><p>· How does your product compare to the competition and do you want to differentiate or blend in?</p><p>· Are your customers, and society at large familiar enough with having the technology in their homes that the embodiment can become more complex?</p><p>Another best practice when designing embodied AI products (and really any physical product) is to tell a Design Story. Every product says something about itself, even ones that don’t actually speak out loud. When a person picks up or looks at a product, they are automatically reminded of something else they have seen or experienced. It’s human nature to categorize things and relate them to past experiences. Acknowledging this allows designers to purposefully and strategically elicit a specific memory or feeling of a specific experience, place, or thing. Are there certain memories you want to trigger, or emotional responses you want to evoke with your AI product? This can be done through the form languages you’re exploring or solely through the Color, Material, and Finish (CMF) you select. CMF elevates the design to make it relevant to cultural trends and emotional expectations. CMF designers mine unspoken global cultural trends and weave that inspiration into the products they design. Combined, the form and CMF tell a metaphorical Design Story. Perhaps you want your AI object to look like a Zen temple, or a high-tech surgical device, or a familiar kitchen utensil. Each of these design stories implies certain decisions on form, color, material, and finish. Here is an example of a moodboard series created to explore these three completely different Design Stories:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZLqokCdDjCHHdIXEx9a_UA.png" /></figure><p>And here is a hypothetical application of these three Design Stories onto a simple form representing an AI-embodied object:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CibKhA5laUtQQ1ygrY1-xA.jpeg" /></figure><p>Hopefully you can clearly see how the initial conceptual Design Stories were fleshed out through moodboards and then translated to CMF decisions in the image above. Now let’s create your Emotional AI Design Stories. Use these questions to guide your process:</p><p>· What experience, place, or thing do you want your AI object to remind people of when they engage with it?</p><p>· Are there certain memories you want to trigger, or emotional responses you want to evoke?</p><p>· Combine the most interesting of the above into 2–3 word combinations such as Serene Precision, or something even more evocative like Mondrian’s Closet, Star Trek Haberdashery, or Huggable House Plant. Then gather images that best capture the combinations.</p><p>From here you can downselect to the moodboard that best represents your product. This moodboard will serve as inspiration for the design team and a North Star vision for the future of the product. The design team would then analyze the chosen moodboard and use it as inspiration as they embark on their design exploration of colors, materials, and finishes to apply to the product.</p><h3><strong>Defining a Clear Hierarchy</strong></h3><p>In Part 1, we looked at hierarchy as an aspect of the Relationship Metaphors we explored. We will return to that once again here, as we recognize how impactful physical form can be on the perception of expectations and the type of relationship that is suggested. Design decisions such as scale, mobility, and physical interaction can dramatically affect how we feel both subconsciously about a product as well as in more practical ways, such as how much control we feel we have over the product. Let’s investigate each of these factors individually:</p><h4><strong>Scale</strong></h4><p>Anyone who’s grown up with or cared for pets knows that the physical size of the animal changes the relationship dynamic. While scale seems to have no implied impact on how much one can love an animal, there is no denying the relationship one has with a pet hamster is quite different than with a Great Dane. Rooted deeply in our evolutionary subconscious, we feel different levels of control over a very small creature which we could easily crush and a large creature which could cause us harm. Their size makes it perfectly clear who is in charge. Similar dynamics are at play with designed things. We have a completely different emotional relationship with our handheld belongings than we do with our cars, or the buildings in which we work. This is due to a subconscious measuring up of our status in relationship to the other thing that is a core part of being human. Upon being presented with an object, some part of our brains asks, “How do I stack up against this thing? Am I more powerful than it or is it more powerful than me?” Social scientists know that we do this instinctively with each other, so it’s only natural to assume we do it with the manufactured world as well. Especially if we begin to consider the manufactured world as possessing qualities similar to a human or an animal. There are clear functional justifications for creating your AI product at a certain scale (portability for example), but I urge you to consider the possible psychological and emotional implications of making your AI product pocket-sized, handheld, human-sized, or house-sized. Does the scale of your embodied AI promote the kind of relationship you want it to have with its owner?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/457/1*lBkkfXapyvlzQW4H6JGEgg.jpeg" /></figure><h4><strong>Mobility</strong></h4><p>Every new parent is intimately familiar with the excitement that accompanies their child’s ability to walk, and with it, the loss of control and anxiety caused by the potential of a runaway baby. There’s no denying the fundamental shift in the relationship between parent and child when this transition occurs. Likewise, the relationship your customers will form with your embodied AI product will differ dramatically based on your decisions around mobility. When it comes to mobility, there are essentially 3 categories into which your embodied AI will fall:</p><p>· <strong>Non-Mobile</strong>: includes things like Alexa which tend to be installed in one place and then are infrequently moved, Jibo who is intended to be hand-carried around the house, and Siri who is taken with us in our pockets all day every day but never moves on her own.</p><p>· <strong>Controlled Mobile:</strong> includes things with wheels like SoftBank’s Pepper, and Asus’ Zenbo which can be commanded to follow you or to go to another room in the house, and Sphero’s BB8 which can be remotely controlled from a smartphone.</p><p>· <strong>Autonomous:</strong> includes things like Roomba, Kuri, and the emerging fleets of self-driving cars by the likes of Waymo, Uber, Tesla, etc.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/920/1*MrvxHcTfBg_-ZBz8SJUHlA.jpeg" /><figcaption>Waymo’s new Jaguar I-PACE vehicle launching with their driverless car service</figcaption></figure><p>If we imagine these three categories on a mobility spectrum from permanently installed to fully autonomous, we can see how a few key relationship factors would likely vary depending on where your product falls. The perceived aliveness, for example, would most likely be higher with a product that can move around on its own. On the flip side, feelings of having control over the product and of the product being dependent on its owner would both likely be higher with a lower mobility product.</p><h4><strong>Physical Interaction</strong></h4><p>On a more practical note, if your embodied AI does not fully rely on voice interaction and has a screen, it must be embedded in such a way that feels both natural and intuitive. Looking at a few examples in the market, we see that it’s difficult to find an ideal location for a screen when the body of the product is at all anthropomorphic. With Jibo, an adorable home companion robot with articulating hips, his entire “face” is a touchscreen, which works beautifully when he is displaying non-interactive digital facial features on it, but it becomes quite awkward when the user has to jab Jibo in the face to select an app or modify settings. Similarly with SoftBank’s Pepper, a humanoid robot popular in Japan, the user must poke and swipe it in its chest to navigate its embedded tablet UI. I believe the lesson here is that if your product has been designed to have distinct components that resemble body parts of any species, avoid also embedding a touchscreen. It will be nearly impossible to find a location for it that feels natural.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/870/1*gyEBSb_Id03qZvlVF_h_Ig.jpeg" /><figcaption>An attendee orders a cup of coffee via the touch screen on a Pepper robot during a media preview of Nestle SA’s human-less cafe in Tokyo’s Harajuku district Thursday. | BLOOMBERG</figcaption></figure><h3><strong>Existential Embodiment</strong></h3><p>When we think about embodiment, the design of the physical form in which the AI is embedded is certainly a logical place to start. But dig a little deeper, and we reveal more philosophical questions about the meaning of intelligence and where it resides. For instance:</p><ul><li>Is the personality we interact with the object itself or does it inhabit the object?</li><li>Does it exist solely inside one object or in the cloud?</li><li>Can it inhabit multiple objects?</li><li>Is the AI personalized to one user or general for any user?</li></ul><p>Cultural context is also a key factor here. Consider, for instance, the differences between America’s Judea-Christian heritage which sees intelligence as residing in “spirit” by which we tend to mean God and human minds, verses Japan’s Shinto / Buddhist tradition which does not see this clear distinction between spirit and body. Generally speaking, in American culture, material things don’t have spirits, souls or intelligence, and when they do, it feels wrong. This influences our thinking and cultural context to feel some suspicion of non-human (and presumably non-god) forms of intelligence. Whereas in Japan, inanimate objects have spirit, and their gods are everywhere — in rocks, trees, rivers and mountains. Unlike our dystopian Blade Runner-esque, Terminator-filled fictional depictions of the future, Japan’s fictional AI future is optimistic, featuring friendly, helpful, companions. While this may not necessarily impact design decisions, it will be important to keep in mind for market adoption.</p><p>To support the kinds of meaningful relationships we’ve discussed in the previous chapters, one might assume that customers would expect to own their AI, and for it to be personalized to them, as opposed to borrowing a universal AI like we do with city bike-sharing today. But as with many design decisions, there are pros and cons. Here are three examples of how ownership could play out in your AI design:</p><h4><strong>Completely Unique</strong></h4><p>Each one starts off and remains completely unique. It personalizes itself in thousands of ways to each customer, resulting in an infinitesimal chance of one AI being the same as another. As a result, the experience of interacting with the AI would feel tailored to you, but ownership and the responsibilities involved with taking care of something would be more heavily weighted.</p><h4><strong>Interchangeable</strong></h4><p>Each one is identical and interchangeable, which enables easy sharing but precludes permanent personalization. In other words, the same code would be deployed to all users and no per-user learning would occur. There would be no memory of preferences or shared memories to harken back on. But in exchange, it would be simple and immediate to start using any version of the AI with the expectation that it would work exactly like every other version you had encountered.</p><h4><strong>Interchangeable + Unique</strong></h4><p>Perhaps the best of both worlds, each one is identical but when someone picks one of them up, it identifies the user and automatically downloads that person’s personalized AI that remembers past conversations and preferences.</p><p>As we’ve seen, our perceptions of AI products and the interactions we have with them are dramatically influenced by their physical form, but just as influential are decisions around the relationship between the intelligence and the object itself. Establishing whether the agent we interact with acts as the soul of the object or a passenger of the object changes the experience of interacting with it entirely. Of similar impact are the architectural arrangement implications of the AI existing only in one object versus many objects through connections to the cloud.</p><h3><strong>He Said, She Said</strong></h3><p>There is a lot being said about gender with respect to AI. Mainly, folks are asking why almost all of today’s virtual assistants have female names and default to female voices. Some theorize that people are conditioned to expect women, not men, to be in assistant roles — and that the [mostly male] makers of digital assistants are influenced by these social expectations. The fact that in Great Britain, Apple launched Siri with a male voice, perhaps because the British have historically had male servants, serves to reinforce the notion that cultural unconscious biases are dictating AI gender decisions.</p><p>Regardless of how we got to where we are today, AI has potential far beyond the virtual assistants available now. As they become smarter, more emotion-aware, and embedded in more capable hardware, so too grows their potential to act more human. And as we know, gender is just one of many ways humans can self-identify. It can be a feature of a person as much as that person wants it to be. When designing your AI embodied object, keep in mind that giving it a specific gender, either through name or voice, or both, will imbue it with meaning based on the history of gender roles of the culture for which you’re designing. Do we even need our virtual beings to have a specific gender in order for us to fall in love with them? Perhaps yours will join the burgeoning movement of gender-neutral AI products being pioneered by the likes of <a href="https://www.equalai.org/">EqualAI</a>, an initiative focused on correcting and preventing gender bias in the development of artificial intelligence.</p><h3><strong>What’s Next?</strong></h3><p>If you’ve read this far in this series, you understand that shifting the way we interact with technology from commands to conversations will reshape our daily experiences with our products. As we develop more meaningful relationships with our possessions they will garner more power, leveling the social playing field, and enabling them to influence our behavior more effectively than ever before. As AI products begin to fill more roles in our lives, designers have an opportunity to carefully craft the relationship we have with them, how they behave, how much control we should have over their personality and behavior, how they reflect the brand that created them, and how the physical embodiment of the AI supports the experience of interacting with it. Hopefully the ideas and frameworks proposed in this series have demonstrated that, if carefully considered, our relationships with AI products can meaningfully enhance our experiences by beginning to resemble our favorite relationships with each other, our pets, and our families.</p><p>Well, that’s it! Thank you for sticking with me on this multi-part adventure. I couldn’t have done it without the support of my many thought partners and editors on this project: first and foremost, my mom and Professor of Psychology at Emory University Dr.Sherryl Goodman, and my fellow <a href="https://www.frogdesign.com/">frogs</a> Jonathan Grossman who produced the mood boards and renders for this chapter and <a href="https://medium.com/@evejweinberg">Eve Weinberg</a> who edited this and all 3 of the chapters, <a href="https://medium.com/u/788e9f70a159">Andrew Haskin</a>, <a href="https://medium.com/u/ca1dba03123">FabioCarnevaleMaffè</a>, Ara Acle, Eli Myers, Christine Todorovich, and Mayaan Albert. Thank you all for your patient editing, research support, inspiration, and advice.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2F8f5f8b%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2F8f5f8b&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href">https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href</a></iframe><figure><a href="https://becominghuman.ai/artificial-intelligence-communities-c305f28e674c"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*2f7OqE2AJK1KSrhkmD9ZMw.png" /></a></figure><figure><a href="https://upscri.be/8f5f8b"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*v-PpfkSWHbvlWWamSVHHWg.png" /></a></figure><figure><a href="https://becominghuman.ai/write-for-us-48270209de63"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*Wt2auqISiEAOZxJ-I7brDQ.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a5dd41a5202f" width="1" height="1" alt=""><hr><p><a href="https://becominghuman.ai/artificial-love-design-for-emotional-intelligence-a5dd41a5202f">Artificial Love: Design for Emotional Intelligence</a> was originally published in <a href="https://becominghuman.ai/">Becoming Human: Artificial Intelligence Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Artificial Love: Design for Emotional Intelligence]]></title>
            <link>https://becominghuman.ai/artificial-love-design-for-emotional-intelligence-9fa2966eb73b?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/9fa2966eb73b</guid>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[emotional-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Tue, 12 Dec 2017 21:00:19 GMT</pubDate>
            <atom:updated>2017-12-13T20:15:35.708Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*I2tI5lqPmiETFpPyml1Wwg.png" /><figcaption>From left to right, <a href="https://www.bostondynamics.com/spot-mini">Boston Dynamics: Spot Mini</a>, <a href="https://www.teenageengineering.com/designs/R">Teenage Engineering: R</a>, <a href="http://consequentialrobotics.com/miro/">Consequential Robotics: Miro</a>, <a href="https://www.woebot.io/">Woebot</a>, <a href="https://www.heykuri.com/">Mayfield Robotics: Kuri</a></figcaption></figure><h4>Part Two: An<strong> Evolving Emotional Connection</strong></h4><p><strong><em>Note:</em></strong><em> This is Part 2 of a multi-part series looking at design opportunities for Emotional AI. Part 1 is </em><a href="https://medium.com/@sethdanielsnyder/artificial-love-designing-for-emotional-intelligence-5b21c11242c8"><strong><em>here</em></strong></a><em>.</em></p><p>Welcome to Part 2 of the Artificial Love series. In Part 1, we discussed the growing shift in the way we interact with technology — from commands to conversations and how that shift will reshape our daily experiences with our products. We argued that designers have the potential to create AI products that we relate to in ways that resemble our favorite relationships with friends, relatives, co-workers, pets, and others. We presented frameworks to help you design products with such relationships in mind, with specific implications for how designers craft the ways that AI products behave, how much control designers have over our products’ personality and behavior. Finally, we proposed a set of aspirational models for the personality of a brand or an AI.</p><p>In this second part of the series, we will provide guidance on how to design products with artificial emotional intelligence that get know us better over time, predict our needs, understand our emotional states, help to alleviate our anxiety, commiserate with us over bad news, and enhance our excitement and joy.</p><h4><strong>Mapping an Evolving Emotional Connection Over Time</strong></h4><p>So how will AI products develop increasingly meaningful connections with us over time? By studying the connections we have with the people we interact with everyday, we can see a familiar theme — developing trust takes time. For example, we rely on our parent’s financial advice when we need their sage guidance, our best friend’s music endorsement when we seek something new to listen to, or our doctor’s instructions to follow a health regimen. These are trusted sources with whom we have well established relationships, based on years of relationship building and experience with their dependability. Or, in the case of the doctor, trust earned based on knowledge of their years of training and expertise. So how will we develop these kinds of trusting relationships with AI products? As our products begin to give us the feeling that they are dependable, understand us, and care for us, we will develop similarly meaningful relationships with them. We will have ever growing expectations of our things and as they fulfill those expectations through perceived sentience, they will inherently garner more influence over our emotions, behaviors, and decisions. As discussed in <a href="https://medium.com/@sethdanielsnyder/artificial-love-designing-for-emotional-intelligence-5b21c11242c8">Part 1</a> of this series, people who live with Alexa and Siri are already reporting feelings of love and trust with their virtual assistants. This trajectory is bound to continue as we live and share experiences with these AI characters and as they increasingly give us the impression that they understand and care for us.</p><p>To help us imagine how deeper connections with our AI products may manifest in the future, let’s look at some of the steps that were likely involved in forming an emotional connection over time with some of our typical relationships, such as with your…</p><p><strong>Mom:</strong></p><ul><li>Step 1: Birth</li><li>Step 2: Experience of being cared for sensitively and reliably</li><li>Step 3: Feeling of being protected and supported no matter what</li><li>Step 4: Countless shared experiences, large and small</li><li>Step 5: Powerful lifelong emotional connection achieved</li></ul><p><strong>Dog:</strong></p><ul><li>Step 1: Spontaneous or intentional decision to adopt</li><li>Step 2: Adoption</li><li>Step 3: Loving parenting/training through early development (housebreaking, etc)</li><li>Step 4: Regular scheduled caretaking (walks, grooming etc)</li><li>Step 5: Powerful emotional connection achieved</li></ul><p><strong>Best Friend:</strong></p><ul><li>Step 1: First introduction</li><li>Step 2: Subsequent positive interactions (Acquiring street cred on the playground, for instance)</li><li>Step 3: Catalytic friend event (something that cements the friendship)</li><li>Step 4: Display of dependability and common interests over time</li><li>Step 5: Powerful emotional connection achieved</li></ul><p>As you can see, the steps involved in forming a powerful emotional connection with another person or thing are highly dependent on what that person or thing is in relation to you. Building trust with a sibling is radically different from doing so with a professor. And one might argue that building meaningful relationships is rarely as simple and straightforward as the linear steps outlined above. From the hormonal desire of a new sexual relationship to the heartbreaking pain of a break up, being in a relationship with another human is infinitely complex. Most of us grow old still being perplexed by it. Many still searching for love, or harboring old grudges with onetime best friends. It is basically a rule of human nature, it seems, that relationships undergo metamorphosis. But it is through those transformations that we intensify the bonds with each other, by learning to understand each other’s changing needs and weaving them in with our own. What then, might a relationship journey look like between an AI product and the person who owns it? Using this Emotional Connection Map as a starting point, ask yourself and your team:</p><ul><li>How will your AI product first recognize its owner? What does their introduction feel like?</li><li>How does the AI product let its owner know it’s listening to them?</li><li>How does the AI product show that it knows what its owner wants?</li><li>How does the AI product show that it knows what its owner is feeling?</li><li>How does the AI product show empathy? How about love?</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*k6NYN9UvJ8qNQhJV1GVehQ.png" /></figure><p>As we progress from the initial introduction with someone, through to them being aware of my interests, there is a lot of nuance to be considered. And things get even more complex if we are lucky enough to push up to the levels of sympathy and empathy with someone. These are design decisions that should go into the development of any new AI product.</p><p>At this point you may be thinking, how can I use frameworks like this if my AI product isn’t yet capable of accurate proactive interactions let alone convincing displays of empathy. I would encourage any product developer thinking this way to envision your current product in the context of an AI technology roadmap such as the one below, and to start making design decisions that set your product on a smooth path towards the future.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rGuBNFTzi1IdqcZYYEgXSA.png" /></figure><p><strong>1. Assistance &amp; Trust</strong></p><p>This is stage 1, and it is pretty much where all AI products are in 2017. During this stage, AI products will build a foundation of reliability and trust with customers as they assist us with daily tasks and seem to understand what we ask of them, at least some of the time.</p><p><strong>2. Prediction &amp; Reliance</strong></p><p>After a few months of shared experience, our AI products theoretically have enough data about us and our individual patterns, that they can be more predictive and proactive — anticipating our needs and desires. When they are able to do this, our relationship will upgrade from trust to reliance. They will influence our behavior in unimaginable ways and even take agency on our behalf. Leveraging their deep knowledge of us, our biometrics, behaviors, and motivation, they will be capable of providing emotional and physical care, perhaps more effectively than our own loved ones. We will truly feel as though these products care deeply about us; our health, our families, our finances, etc. even though they will not yet actually be capable of caring emotionally in the human sense of the concept.</p><p><strong>3. Empathy &amp; Love</strong></p><p>Finally (some scientists predict as soon as 2040), our AI products will develop convincing empathy and be able to feel emotions, either human-like or wholly different. In the first half of this stage, conversational UI will pass the Turing Test, proving to be so convincingly real, it will be indistinguishable from a human (think, Scarlett Johansson’s character in the movie Her). But as impressive as that will be, it will still be just a clever trick. Only in the latter half of this stage will AIs develop the ability to not only express emotions but to truly feel them. Nobody knows exactly how this will be accomplished and what the results will be once it is, but there is no doubt it will utterly transform our relationships with our products.</p><h4><strong>Interaction Models for Connection</strong></h4><p>Taking your Emotional Connection Map one step further now, let’s look at some potential Interaction Models for developing emotional connections over time. Does your AI product get to know its owner progressively over time through passive observation and shared experiences? Does it ask its owner what their needs are up front in an initial getting-to-know-you interview? Does it try to make suggestions based on rough guesses and use trial and error, thereby learning to refine its understanding? Does it perhaps go through its own developmental stages, i.e. infancy, toddlerhood, etc., and heighten its ability to understand its owner with each progressive stage of growth? Or is it some combination of all of these? The following charts depict these discrete opportunities to explore when designing your embodied AI experience:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cgFYZJ1mJYJSb0nwfpKOyQ.png" /></figure><p><strong>Conversation &amp; Shared Experience: </strong>The AI gets to know its owner through natural conversation over time as well as having shared experiences and observing its owners’ reactions to life events. The potential limit of how well the AI can know its owner is capped entirely by the technology capability. An assumption is that there is a minimal default understanding established during product setup by gathering personal data from various sources.</p><p><strong>Data Input: </strong>The AI gets to know its owner through either a one-time interview-like conversation or through manual data input of preferences. This way the owner has control over how much (and how) the AI can know them.</p><p><strong>Trial and Error: </strong>Starting from scratch, as if complete strangers, the AI gets to know its owner through conversational trial and error. It makes guesses and suggestions and learns from each interaction, sometimes resulting in confusion, but over time, coming to clarity.</p><p><strong>Developmental Stages: </strong>The AI goes through its own developmental stages from infancy to youth to adulthood, at each stage of which its ability to understand its owner reaches a limit and flattens out, until the next stage is unlocked.</p><h4><strong>Learning Methods for AI Products</strong></h4><p>So far in this chapter, we’ve looked at methods for developing emotional connection over time. Here we will look at the functional side of how AIs can learn, that will support those relationships. Here are three categories of education methods to be explored with your AI product:</p><p><strong>Direct Interaction</strong></p><ul><li>Learns by what you ask of it (task-based)</li><li>Learns by what you do with it (physical interaction)</li><li>Learns by what you share with it (conversation-based)</li></ul><p><strong>Passive Observation</strong></p><ul><li>Learns by observing interactions of you with others</li><li>Learns by observing your behaviors or reactions</li><li>Learns by observing and detecting qualities of its environment</li><li>Learns by inferring your mood through your tone of voice and the words you use</li></ul><p><strong>Active Investigation</strong></p><ul><li>Learns by mining/integrating with other services and online profiles</li><li>Learns by gaining insights from a central intelligence</li><li>Learns by talking to other AIs</li></ul><p>The most likely scenario is that your AI product will use a combination of these approaches to learn about its environment, understand your context, and, based on these, react to you appropriately and intelligently. Data scientists call this phenomenon “Context Fusion” and it’s essential to making AIs seem human-like, seeing as how humans perform Context Fusion far better than any existing computer algorithms. Soon though, AIs will have two significant advantages over humans: infinite memory and global networking. Among the many superpowers endowed on them by their makers who stuff them with sensors, cameras, and data processing units, infinite memory and global networking will enable their Context Fusion to achieve levels of understanding beyond what humans are capable of. But that’s a conversation for another day.</p><h4><strong>Communication Methods</strong></h4><p>It’s well known that communication is the foundation of any successful relationship, be it personal or professional. So now that we have an AI product with a distinct Relationship Metaphor, a well-considered personality, a model for getting to know its owner(s) over time, and a powerful set of sensing superpowers with which it can perform heroic Context Fusion, the next step in the design process is to define how users will communicate with it and how it communicates with them. But how will you choose which communication methods are right for your AI embodied product? To help answer this question, let’s first think about both sides of the communication equation: inputs (human to AI) and outputs (AI to human). Depending on the hardware payload and software capabilities of your product, there may be many options for ways to communicate in both directions with the AI; common examples today being voice, non-verbal sounds, eye contact, and gesture. Amazon Echo’s communication interaction model, for example, starts with a verbal wake word and then a command phrase that triggers the Echo to display lights (presumably to indicate that Alexa is listening) followed by a verbal response and, if all goes well, successful execution of the command. SoftBank’s Pepper robot, by contrast, can track your gaze, make eye contact, read your emotions, respond verbally, and make physical gestures carried out by its vaguely humanoid upper body.</p><p>Starting with this (admittedly non-exhaustive)<strong> </strong>list of potential inputs and outputs, use the following framework to down-select to the most appropriate communication methods:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AgQJCn4W8TBLZ05etJpW3g.png" /></figure><p><strong>Capabilities: </strong>Start by cutting the list down to what your device is capable of in terms of inputs and outputs.</p><p><strong>Personality: </strong>Use the Relationship Metaphor and Behavior Model you developed from <a href="https://medium.com/@sethdanielsnyder/artificial-love-designing-for-emotional-intelligence-5b21c11242c8">Part 1</a> to inform the communication methods. Are there certain inputs and outputs that are intrinsic to the personality of the AI you’re crafting? If you want your product to behave motherly, for instance, you might consider how touch and body language could enhance the caring and protective relationship you want it to have with your users.</p><p><strong>Context: </strong>Determine how users will be engaging with the AI embodied object and ask if there are certain inputs and outputs that are intrinsic to that context. For instance, if the AI is in a car, safety and intrusiveness are primary concerns so text and voice outputs may not be appropriate.</p><p><strong>User Intent: </strong>Perhaps you want to implement distinct communication methods based on the type of communication the user initiates. For instance, if the user commands the device to do something, it may be more appropriate to just initiate the action and communicate completion with light or sound, whereas if the user initiates a conversational dialogue, it may feel most natural for the device to respond in kind with voice and gesture.</p><p><strong>User Preference: </strong>Do you want users to be able to customize how they communicate with your product? Apple’s Notification Center is an example of how much some users love customizing how their iDevices communicate with them, but this can be a slippery slope of scope creep.</p><h4>Are Current Voice UIs Skeuomorphic?</h4><p>When it comes to selecting communication methods, it’s important to recognize that we are still in very early days of what promises to be a long and investment-rich development of human-robot interaction. Some even compare the current state of Voice UI to the skeuomorphic digital design that pervaded screen-based experiences for decades and which has since been replaced with flatter, simpler interfaces. The value of skeuomorphism in digital design of course, was widely recognized as creating a bridge between the physical world and the brand new digital world, helping users understand functionality and intent. For users who had never encountered a digital volume knob, for instance, it made sense to design the two-dimensional screen knob to look as much like a physical knob as the pixels would allow, giving those users an automatic sense of knowing how to interact with it. As society at large became more comfortable with digital-first experiences, interactions and visual designs were able to advance to a way of communicating that feels like a more natural expression of the pixel-bound medium. Where then does that leave us with designing how we communicate with our AI products? Should they attempt to sound like a human, just as digital buttons in the early 2000’s resembled their physical counterparts? The primary advantage of enabling human-like voice, similar to digital buttons that resemble their physical counterparts, is to aid in discovery (most Siri and Alexa users can only remember a small fraction of the commands the AIs are capable of). Once users discover that they can speak to their AI products as if they are human, they can more easily learn about the capabilities and benefits of the product. But, the less often discussed benefit, is that voice can go a long way to support the kinds of emotional connection discussed in <a href="https://medium.com/@sethdanielsnyder/artificial-love-designing-for-emotional-intelligence-5b21c11242c8">Part 1</a>. Many thought leaders are theorizing that voice assistants and other AI products that are programmed to sound human will be replaced with a less skeuomorphic communication medium that has yet to be defined. Perhaps your AI product will be the first to cross this chasm, to communicate in a way that feels completely natural for the AI medium, while enabling deep emotional bonds to form (R2D2 anyone?).</p><h4>What’s Next?</h4><p>In this second part of the Artificial Love series, we built on the premise that our evolving emotional connections with others over time are contingent on increasingly better understanding the other’s needs. We investigated how those connections may look when one member of the relationship is an emotionally intelligent AI product, such as a chatbot, robot, or autonomous car, just to name a few. It may be hard to imagine it today, but one day in the not too distant future, super-sophisticated intelligent AI beings will develop complex bonds with humans. What doesn’t take imagination is the fact that already today, we are seeing people form unidirectional relationships with their virtual assistants like Siri, and with robots like the Roomba. Despite the fact that the feelings aren’t yet able to be reciprocated, we can already see the power these relationships can have over the people in whose pockets, desktops, homes, and cars they live.</p><p>Stay tuned for the next and final installment in this series which will look at the critical decisions to be made around embodiment — the size, shape, and physical characteristics of your AI product. We’ll also discuss the opportunities to explore embodiment mechanisms such as the AI living solely inside one object, versus in the cloud, accessible from anywhere. Finally, we will address the elephant in the room — gender. I hope you come back to check it out and join in the discussion!</p><p>I’d like to thank my thought partners on this project: first and foremost, my mom and Professor of Psychology at Emory University Dr.Sherryl Goodman, and my fellow <a href="https://www.frogdesign.com/">frog</a> <a href="https://medium.com/u/1e72b0be262b">Eve Weinberg</a>. Thank you for your patient editing, research support, inspiration, and advice.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2F8f5f8b%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2F8f5f8b&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href">https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href</a></iframe><figure><a href="https://becominghuman.ai/artificial-intelligence-communities-c305f28e674c"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*2f7OqE2AJK1KSrhkmD9ZMw.png" /></a></figure><figure><a href="https://upscri.be/8f5f8b"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*v-PpfkSWHbvlWWamSVHHWg.png" /></a></figure><figure><a href="https://becominghuman.ai/write-for-us-48270209de63"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*Wt2auqISiEAOZxJ-I7brDQ.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9fa2966eb73b" width="1" height="1" alt=""><hr><p><a href="https://becominghuman.ai/artificial-love-design-for-emotional-intelligence-9fa2966eb73b">Artificial Love: Design for Emotional Intelligence</a> was originally published in <a href="https://becominghuman.ai/">Becoming Human: Artificial Intelligence Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Designing Mythologies:]]></title>
            <link>https://medium.com/@sethdanielsnyder/designing-mythologies-9716269a52ed?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/9716269a52ed</guid>
            <category><![CDATA[design]]></category>
            <category><![CDATA[mythology]]></category>
            <category><![CDATA[star-wars]]></category>
            <category><![CDATA[future]]></category>
            <category><![CDATA[film]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Mon, 28 Aug 2017 00:01:02 GMT</pubDate>
            <atom:updated>2017-08-28T00:01:02.441Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ijl8EnbIUpJyFqL6HoCrlQ.png" /></figure><h4>Building Worlds to Frame the Future</h4><p>From the first time we saw a lightsaber slice through the air with the full power of the Force behind it in Star Wars, any stick or plastic sword has had the potential to take on that electric schvrmmmmmmm sound. Watching Joaquin Phoenix’s character fall in love with his Artificial Intelligence system in the movie Her made many of us think about our relationship with Siri or Alexa a little differently. Good storytellers, especially in science fiction and fantasy, build worlds with the power to give new meaning to our perceived realities. The mythologies created for these fictional worlds often take what we know to be true, then extrapolate by incorporating cutting-edge science and experimenting with new rules for physics, government, anatomy and religion. Designing these mythologies is good entertainment, but it can also be a powerful design tool. As a framework, designing for mythologies gives designers a new way to envision and design for the future.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*fjks9EVwD0UJQ-ZezpAEuQ.gif" /></figure><p><strong>Real rules for fantasy worlds</strong></p><p><a href="https://vimeo.com/groups/183185/videos/38026023">Director George Lucas described</a> creating the vast world of Star Wars by saying, “The whole culture has to be designed. What do they believe in? How do they operate? What are the economics of the culture? Most of it doesn’t appear in the movie, but you have to have thought it through otherwise something always rings very untrue or phony about what’s going on.” Thus, designing mythologies requires grounding fictional scenarios with some rules to enable stories to exist within a richly developed, holistically considered world. Recalling his decisions to put rules around materials that <a href="http://starwars.wikia.com/wiki/Lightsaber#Lightsaber">lightsabers</a> could and could not cut through, Lucas adds, “One of the things I try to achieve is a sort of immaculate realism in a totally unreal and fantasy world. I can make up the rules, but once I make it up, I have to live with it.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ch0dPk116rz6Kjq64L9l7A.jpeg" /></figure><p>Working closely with directors, production designers create and develop the overall visual look, atmosphere and mood to support a story. Through location, costumes, set design and props, they establish the rules that govern the story’s world. The Batmobile, Harry Potter’s wand, the aliens in Men in Black, TARS in Interstellar, the earplugs in Her — these were all overseen by the production designer to support the mythology of their respective worlds.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/380/1*DbntjGHh5_YScBAqG4PplA.gif" /></figure><p>These fantastic worlds are entertaining, but they also help us understand, address and more responsibly design our own future. For example, watching The Jetsons might help us envision a flying car. To actually design this car, however, we would have to establish the larger world of the flying car. We would need to consider the future driving test, road signage, legal and ethical standards and other considerations to make flying cars feel real and achievable. To do this, we would need to reach into the future and seriously consider the desires and pain points of key members of that world, from the taxi driver whose flying car breaks down during his shift to the politician working to loosen regulations on who can operate flying cars. In other words, we must create complete mythologies for the things we design and the worlds in which they exist.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*MdwjEM7ysayqdNeGIdKrpA.gif" /></figure><p><strong>Designing mythologies for business success</strong></p><p>Similar to film production designers, frog designers have a process called “Futurecasting” that we use to design products and services for new worlds in partnership with our clients. <a href="https://www.fastcodesign.com/3065080/frogs-5-steps-to-predicting-the-future">Futurecasting</a> is grounded in scenario planning, in which we craft possible futures as a means of understanding what lies ahead for an organization. This exercise produces valuable insight that informs decision-making, such as necessary investments in technologies to pursue or skills to acquire. Examining these futures also sparks associations and creates momentum for suggesting potential behaviors that will lead towards a desired future.</p><p>Over several years, frog partnered with the semiconductor company <a href="https://www.frogdesign.com/work/rambus-iot-vision-prototypes">Rambus</a> to develop a new IoT strategy and a series of emerging technology prototypes. To help them cultivate a new vision, we used Futurecasting to bring it to life through design and film. To do this, the team ran several collaborative workshops to design mythologies for a future in which every object on Earth is connected to the internet. The results of these sessions included dozens of product and service concepts, and business opportunities for Rambus to own their industry in this possible future. This led Rambus to experiment with a completely new sensor technology, essentially reorienting the organization’s trajectory.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplayer.vimeo.com%2Fvideo%2F199244878&amp;url=https%3A%2F%2Fvimeo.com%2F199244878&amp;image=https%3A%2F%2Fi.vimeocdn.com%2Fvideo%2F612441725_1280.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=vimeo" width="1280" height="720" frameborder="0" scrolling="no"><a href="https://medium.com/media/98fb5456908a20492c364d5cbbcd1e15/href">https://medium.com/media/98fb5456908a20492c364d5cbbcd1e15/href</a></iframe><p>Much like ancient mythologists used storytelling to communicate the unknown, designers need to communicate humankind’s future experiences to help build shared meaning. Our goal as designers is to responsibly consider societal implications as an input into the design process. We do this to weave a grand narrative tapestry of a culture, composed of the human stories of the technological advancements that have enabled future products and experiences.</p><p><strong>Why designers should be mythologists</strong></p><p>It is already known that the better we understand the people we design for, the more successful our new products can be. I propose a new role for designers, which is that of the mythologist. This would involve writing stories, customs, laws and traditions. Designing paradigms for the future worlds that new products or experiences will inhabit. As Lucas did when creating the mythology of Star Wars, design mythologists define these new worlds in detail beyond what is needed to simply communicate product features. We will push beyond flat archetypes and made-up personas — where a lot of design research stops — to establish the rules and context in which people would actually live with our designs.</p><p>Designing mythologies to support products and experiences is a lens for designers to create a more complete, higher resolution picture of the world in which their users live. These mythologies are expressions of psychological, cultural and societal truths. Empowered with such truths, designers can create more thoughtful and impactful things that reflect the holistic context customers live in, not just their specific functional or emotional needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*s7F71TyG-HNzWTechoqa4Q.gif" /></figure><p>Every day as designers, we are contributing to creating a new world — one filled with international social networks, virtual personal assistants, and self-driving car fleets. In lieu of a long history of shared culture with these technologies, we need new stories and mythologies to address the societal implications of these transformations. In the years to come, our global society will face perpetual disruption at unprecedented scales in the way we manage information, communicate with each other and deliver goods. We will continue to see the disruption of centuries-old institutions that define our civilization, within sectors like education, government and healthcare. As we transform these systems, we as design mythologists can examine what they do, what they mean to us, and how they fit into our personal and shared narratives of what the world is, and what it should be.</p><p>The role of the designer is always evolving, but a remaining constant is that clients will continue to ask designers to make the future a reality. As long as we are projecting into the future, let’s send into that infinitely vast fourth dimension not only exceptional design, but also immaculately realistic mythologies of the not yet.</p><p>This article was originally published on <a href="http://designmind.frogdesign.com/2017/08/designing-mythologies-frame-the-future/">frog’s DesignMind</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9716269a52ed" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Artificial Love: Design for Emotional Intelligence]]></title>
            <link>https://becominghuman.ai/artificial-love-designing-for-emotional-intelligence-5b21c11242c8?source=rss-7955690dd2ca------2</link>
            <guid isPermaLink="false">https://medium.com/p/5b21c11242c8</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[emotional-intelligence]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[robotics]]></category>
            <dc:creator><![CDATA[Seth Snyder]]></dc:creator>
            <pubDate>Fri, 14 Jul 2017 04:02:47 GMT</pubDate>
            <atom:updated>2017-12-16T06:20:35.835Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TZQgVxlneYtKXrEX9nxDsw.png" /><figcaption>Zenbo, Waymo, Sophia, Elli-Q, and Azuma Hikari</figcaption></figure><h4><strong>Part One: In Search of Meaningful Relationships</strong></h4><p><strong><em>Note:</em></strong><em> This is Part 1of a multi-part series looking at design opportunities for Emotional AI. Part 2 is </em><a href="https://medium.com/@sethdanielsnyder/artificial-love-design-for-emotional-intelligence-9fa2966eb73b"><strong><em>here</em></strong></a><em>.</em></p><p>We all have a relationship with the things we own. We grumble at our alarm clock for waking us up out of a dreamy sleep, we give our cars names and talk to them when they make funny noises. But what happens when they talk back? When our things have the ability to clearly communicate with us, to understand us, and to express realistic emotion, the depth of those relationships will grow exponentially. In this emergent age of artificial intelligence, conversational UI, digital personal assistants, and autonomous everythings, there is an opportunity for designers to craft the roles these emotionally intelligent products play in our lives.</p><p>As machine learning capabilities advance, users will feel better understood and, will naturally, grow closer to their AI products. This begs the question: what kind of relationship do we want to have with our AIs? And for designers and product developers a whole Pandora’s Box of questions arises: what kind of relationship would bring the promise of your product to life through artificial emotional intelligence? If your product behaved more like a friend, how would that influence the user experience? What if it had needs of its own, maybe even needing care and love itself? It could perhaps be more like a pet and maybe not communicate with words at all, but rather with movement and sounds like R2D2. This is Part 1 of a multi-part series that will outline an approach to designing emotionally intelligent AI products that not only play a meaningful role in our lives but also redefine our relationship with technology.</p><h3><strong>How We Got Here</strong></h3><p>To help us understand why these questions are especially challenging and important now, let us take a quick trip back through a vastly simplified version of the history of product design.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JXfM1R5pn-BTsOo6n9XuMg.png" /></figure><p>First, there was the design of physical objects, such as tools, furniture, vehicles and signage. These were objects a user could directly observe or interact with to intuitively understand their functionality.</p><p>Along came the first computers and their arcane software. These early text-based digital experiences required a massive learning curve, limiting usage to experts in laboratories. Incidentally, it was around this time that the mathematician Alan Turing invented the idea of intelligent machines, a concept he deemed to be absurd. Instead, he saw the potential for computers (which were the size of an entire room at this point) to trick humans into thinking they were human; hence he devised what came to be known as the Turing Test — a social experiment to determine if a given software could successfully pose as a human.</p><p>Eventually, friendlier graphical interfaces were created to help everyday people interact with computers and design paradigms have developed over time. Even still, digital interfaces remain complex, with features hidden behind menus and tasks requiring many clicks through clunky flows. The entire field of Interaction Design emerged to take on this challenge, to help us navigate an application and complete an objective easily and without error. And in some ideal cases, to elicit an emotional response from the user, further enhancing the experience and the connection to the brand.</p><p>Finally, we arrive at the artificially intelligent, often screenless “interfaces” of today (examples pictured below). Much like the early Wild West days of computers and the web, there are no standards or best practices. Yet, unlike the history of software, the technology has skipped past limited usage by expert users, leaping right into the homes and pockets of consumers. While these interfaces open up a brand new set of UX problems by virtue of being invisible, or at best, abstract, the potential for building trusting, meaningful relationships with customers is huge. Even though most AI products are still probably a long way from passing the Turing Test, we are seeing people form emotional bonds with them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*rinikw-jY5IzlUb5fzLvMQ.gif" /></figure><p>And this is just the beginning. The technology available today is a mere beacon of things to come — akin to primordial humans. As the technology advances from the single task Artificial Narrow Intelligence of today (IBM’s Deep Blue chess AI, for example) towards Artificial General Intelligence (AIs that are as smart as humans), the associated design and ethical challenges become even more critical. But along with the challenges, come significant opportunities. Human-level intelligence will mean our products will be able to read our facial expressions, answer abstract questions, creatively express themselves, and many other as yet undetermined abilities. We will finally feel as though our products actually understand us, as humans understand each other, unlocking not just practical abilities like anticipating our needs, but also the emotional support that comes with meaningful human relationships.</p><h3>From Mistrust to Love</h3><p>For these kinds of human-machine relationships to flourish, trust is a prerequisite. Creators of AI products will need to alleviate consumer’s privacy concerns and demonstrate responsibility with personal data before dependency can grow and deeper relationships can emerge. It can be assumed that for many, the level of intimacy implied by living with an emotionally intelligent AI product equates to <a href="https://www.theatlantic.com/technology/archive/2016/05/the-privacy-problem-with-digital-assistants/483950/">an infringement of privacy</a> that is unacceptable, at least as long as personal data is being sent back and forth on the internet. Local AI neural network technologies like <a href="http://news.mit.edu/2016/neural-chip-artificial-intelligence-mobile-devices-0203">this one being developed at MIT</a> could help, But many believe that as the technology continues to drive more realistic personalities that deliver more value and familiarity with the technology grows, consumer anxiety will fade to the background. As an early indicator of this trend, <a href="https://www.jwtintelligence.com/trend-reports/speak-easy/">a research study</a> conducted by J. Walter Thompson Innovation Group London and Mindshare Futures, in which 100 consumers were given Amazon Echos for the first time, found that,</p><blockquote>“the emotional response to Alexa grew during the course of our experiment as people became more comfortable using it, pointing to the latent potential for a closer relationship.”</blockquote><p>The study also included over a thousand UK smartphone voice assistant users,</p><blockquote>“for whom a deeper emotional attachment is starting to develop. Over a third of regular voice technology users say they love their voice assistant so much they wish it were a real person. Even more astonishing is that more than a quarter of regular voice technology users say they have had a sexual fantasy about their voice assistant.”</blockquote><p>For creators of AI products, the challenge will be to navigate these ever-closer relationships as the technology becomes increasingly capable of understanding us and consumer mistrust gives way to love.</p><h3>Relationship Metaphor</h3><p>In the years to come, our global society will face perpetual disruption at unprecedented scales as a result of Artificial Intelligence. Disruption of the way we communicate with each other and our things but also of the centuries-old institutions that define our civilization: education, government, healthcare, and work. As we transform these systems, by giving our cars autonomy or democratizing education through robotics, we’re not necessarily changing what these systems do, e.g. drive, teach, but what they mean to us, and how they fit into our personal and shared narratives of the world and our relationships within it. While much of this future may seem scary, at least partly for being unknown, what we do know is that people are more comfortable with new technologies when they have a clear understanding of the role it will play in their lives. Defining the kind of relationship you want people to have with your AI embodied object is fundamental to the experience. It sets expectations, defines behavior, and goes a long way toward making people feel comfortable with what could otherwise be a creepy intrusion to their lives. To do this, we need to establish the right Relationship Metaphor — a framework that allows people to compare their relationship with the AI with that of a familiar relationship in their life.</p><p>To take steps toward selecting and, subsequently, refining the right Relationship Metaphor, consider some stereotypical relationships such as those listed below. First, try matching up the core functionality of the AI in your product (e.g., autonomous car, virtual personal assistant, robot companion) to this list of common relationship metaphors (or create a new one). See if any feel like a natural fit, and, once you’ve identified the best fit, also push yourself to imagine what the other relationships might imply for your product.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*poN96QwBGcNdXOQUExegtg.png" /><figcaption>— — — — — — — — — —</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8cN0Ti1kJwi6VKSURVjWXg.png" /><figcaption>— — — — — — — — — —</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NSjp6lXVtsUTnJkMl9aTyg.png" /><figcaption>— — — — — — — — — —</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PerXgdXWF71qTmreD3OANw.png" /></figure><h3><strong>A Platform for Ideation</strong></h3><p>As you work your way through each Relationship Metaphor, use it to guide concept development. Ask yourself and your team:</p><ul><li><em>How would this metaphor be reflected in how the user interacts with the product?</em></li><li><em>What features would be emphasized for each metaphor?</em></li><li><em>How would it communicate it’s intended role in its owner’s life?</em></li><li><em>How would its communication style be different for each metaphor?</em></li></ul><p>As an example, if an autonomous car took on the best friend Relationship Metaphor, perhaps the car would teach you how to fix itself or coach you on how to drive more safely. Maybe it would be able to recognize each family member in the car and communicate accordingly. What if it knew you were a car buff, so it kept you company on road trips telling you about the cool new features in cars as they passed you on the highway? The Relationship Metaphor is a fruitful jumping off point, but will also serve as a North Star vision guiding ideation and decision-making throughout the design and development of your AI-embedded product as well as in your evaluation (and further refinement) of the product. It can help filter out ideas or features that may seem valuable, but do not contribute to a cohesive personality that sets clear expectations and establishes a specific role in your customers’ lives. It also serves as a platform from which the personality of your AI can be further refined.</p><h3><strong>Modeling a Personality</strong></h3><p>Drilling one level deeper, most psychologists define personality as the individual differences in characteristic patterns of thinking, feeling and behaving that come together as a whole to define a person. If we think of personality as the building blocks of who we are, we can begin to imagine designing the nuances of those building blocks for our AI personalities. Thanks to psychology research, work place consulting strategy, and fictional worlds portrayed in videogames, movies, and tv, there are a multitude of examples of frameworks to use to model personalities. Here I’ll highlight four examples that each push the AI designer to think differently about the personality they’re crafting.</p><ul><li><strong>The Five Factor Model, otherwise known by the acronym OCEAN: </strong>A commonly referenced and empirical framework used in psychology that describes five main building blocks that can lay the foundation for other more contextual and specific traits. The five factors are: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism and they each exist as a spectrum from low to high, as shown in the below image. As you begin to imagine where you might want to map your new AI personality, keep in mind that many psychologists negatively judge the low end of conscientiousness and the high end of neuroticism, as they can sometimes be indicators of mental illness or challenges to interpersonal relationships. Whereas both ends of the other 3 factors are considered neutral and variety is seen as valuable within individuals and across relationships.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mWWmIWiTczVFGycMJhP_Ow.png" /></figure><ul><li><strong>Business Chemistry: </strong>While designed primarily to help businesses build successful teams, this framework developed by Deloitte, categorizes everyone into four scientifically based patterns of behavior. Think of this framework as a way to create the perfect relationship pair between a person and their AI. As seen in the below image, the circle is split into four quadrants that make up each behavior pattern, but interesting combinations emerge at the intersections of the four main categories.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3lWg7vbwkzfE2FoOl5Cd_Q.png" /></figure><ul><li><strong>Westworld Attribute Matrix: </strong>Taken from the popular HBO show featuring a theme park of sorts inhabited by robotic AIs, this framework is designed to give the creators of the robots the ability to manipulate the detailed personality traits and mental capabilities of each robot. Imagine if your consumers could have this level of control over their AI product.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QrMbLeSUzXfFBQf28KG9aw.png" /></figure><ul><li><strong>The Sims Personality Points: </strong>Taken from the setup experience in the life simulation game, this framework gives players 25 Personality Points to spend on any of 5 categories, thereby forcing them to prioritize and create a more balanced Sim. If given the opportunity to customize their AI, would consumers benefit from having these kinds of constraints?</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/808/1*z-Duup4Ms6_lI5d8xz963w.png" /></figure><h3><strong>AI as Brand Persona</strong></h3><p>These personality frameworks, along with many other examples not detailed here, such as Myers-Briggs, Enneagram, and the CARE profiles, can help us define with greater detail, how our future AIs will behave and the role they will fill in our lives. But stepping back a bit, so far we’ve been looking at shaping the personality of a single being; but what about the company providing these beings? Lest we forget that these AI personalities will be walking, talking, blinking, thinking representations of the brand that creates them. Much like a customer service professional trained to help customers in a manner that fits with the communication style and values of the brand they represent, so too should your AI personality embody your brand. Marketers and brand strategists will tell you that every company has a voice and a personality with which they express their core beliefs. So what is yours? And how should it translate into the personality and behavior of your AI product?</p><p>Let’s start with the notion that as consumers, we all find ourselves being drawn to certain brands more than others. Branding professionals will tell you that is because we have a personal connection with those brands’ archetypes — the personification of those brands on a human level. We’re subconsciously enamored by Coke’s nostalgia, we’re inspired to explore the outdoors with Patagonia, we feel the love with Subaru. Building on what the brand stands for, archetypes borrow well-known character types from history, film, literature, and pop culture to define a brand’s underlying personality and character. Consider Serena William’s strength and determination and Elon Musk’s inventiveness. Think about Indiana Jones’s carefree adventurousness or Hermione Granger’s encyclopedic knowledge and diligence. Archetypes like these represent the essence of who we are and who we aspire to be, and they can serve as aspirational models for the personality of a brand or an AI. Swiss psychiatrist Carl Jung introduced the concept of archetypes over 50 years ago as universally understood characters that provide emotional prompts to fundamental human desires. Today many companies use an adapted set of Jung’s archetypes (seen in the framework below) to help them determine their brand’s true identity — the character their brand is meant to take on.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rqnif5RrimP82QxyYl0HPw.png" /></figure><p>Now we have an opportunity to adapt these archetypes to shape the personality of AI products and to use them to suggest the kind of relationship these products can have with their owners. Much like with the Relationship Metaphors discussed earlier, Brand Archetypes can serve as a guiding vision not only for your AI personality, but for all aspects and touchpoints of your brand. Once you’ve determined where your brand fits in this archetype framework, use these questions to guide your ideation process:</p><ul><li><em>How would this archetype be reflected in how the user interacts with the product?</em></li><li><em>What features would be emphasized for this archetype?</em></li><li><em>How would its communication style be different as a result of the archetype?</em></li><li><em>What kind of relationship would your customers have with this archetype?</em></li></ul><h3><strong>Behavior over Time</strong></h3><p>We know from psychology research that human personalities are a product of our DNA passed along to us over the roughly 300,000 year history of homo sapiens, and that we acquire directly from our biological parents, combined with our environment. Nature and nurture. All humans have a range of personal qualities that are fluid across situations and over time, yet we strive to categorize each other into personality buckets — “He’s funny” “She’s introverted”, “He’s a ENFJ” etc. When it comes to designing AI personalities, we can choose to replicate the way human personalities develop and are described, or we can try something new. Let’s call this the Behavior Model. How would we characterize the personal qualities of our AI? Are they fixed, do they change on their own (“learning” from experience), or can they be manually adjusted? Here are some examples of ways to frame a Behavior Model for your new AI:</p><p><strong>Singular Growth</strong></p><ul><li>A single personality that is highly responsive to its early experiences with you, mimicking your behaviors and learning from your reactions.</li><li>It decides whether to be a mom, a dog, or a coach based on initial interactions with you and then is that personality forever.</li><li>Imagine an in-home health robot designed to motivate you to change your unhealthy habits. You bring it home and it begins asking you questions to get to know you better. At first, it sounds a bit generic, but over time, it develops a distinct motherly tone. It’s tough on you and always seems to know what’s best for you. It can be a bit overbearing sometimes but you appreciate it because you know it truly cares for you.</li></ul><p><strong>Elastic Adaptation</strong></p><ul><li>A single personality that shifts towards other personalities based on context.</li><li>It can detect your physical and emotional context (i.e. work, out on a date, feeling sad at home) and adapt its interaction style to be most appropriate in that set of circumstances).</li><li>Imagine another health robot. This one starts off sounding a bit like a doctor — straightforward and scientific, but then you notice that it changes its communication style depending on how you’re feeling and what you’re doing. After a tough day at work, it tries to cheer you up like a best friend would. When you have your boyfriend over for dinner, it becomes more submissive and loving, like a pet.</li></ul><p><strong>Fragmented Collective</strong></p><ul><li>Multiple distinct personal qualities coexist and emerge on demand when you call on one of them. You can also manually customize the intensity of those qualities.</li><li>It can completely switch personalities between Mom, Dog, Sidekick, etc depending on which one you want to interact with. You can turn down the intensity of tough love in mom-mode, for example.</li><li>With this health robot, you have become accustomed to manually controlling its personalities. You prefer the coach personality because it’s encouraging and helps you remember to take your medicine, so you usually keep it on that one. Sometimes it’s a bit overzealous with motivation, so you tell it to turn down its intensity. Other times, when you just need some emotional support, you tell it to switch to the dog personality.</li></ul><p><strong>Fixed Family</strong></p><ul><li>Personalities come hard-wired into each of many available versions.</li><li>You can choose to buy multiple copies of the product if you want a few personalities. You may want to own both a Best Friend and a Mom, for instance.</li><li>You have two copies of this health robot — a doctor personality at home, where you feel most comfortable discussing your medical information, and a dog version at work that gently reminds you to take your medicine throughout the day, while providing emotional support.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zmp8HjNDVQ5QXjIyqGkvpg.png" /></figure><p>Mapping these 4 Behavior Models on this graph shows a range from models with little to no change over time to those with high change over time. As well as models that proactively change attributes of their personality and those that do so reactively. Use these questions to guide your ideation process:</p><ul><li><em>What would it mean for your product if it’s personal qualities were proactive versus reactive?</em></li><li><em>What would it mean for your product if it’s personal qualities were dynamic versus static?</em></li><li><em>How do you imagine the most common use cases of your product would be different for each Behavior Model?</em></li><li><em>What are the potential benefits and drawbacks of your product with each Behavior Model?</em></li><li><em>Are any of these Behavior Models are a natural fit for your product? If not, try creating a new one.</em></li></ul><h3><strong>In Search of Meaningful Relationships</strong></h3><p>Shifting the way we interact with technology from commands to conversations will reshape our daily experiences with our products. As we develop more meaningful relationships with our things, they will garner more power, leveling the social playing field, and enabling them to influence our behavior more effectively than ever before. We don’t think twice about following our parent’s financial advice when we need their sage guidance, our best friend’s music endorsement when we need something new to listen to, or our doctor’s instructions to follow a health regimen. These are trusted sources with whom we have well established relationships. As AI products begin to fill these roles in our lives, designers have an opportunity to carefully craft the relationship we have with them, how they behave, how much control we should have over their personality and behavior, and how they reflect the brand that created them. Hopefully the frameworks proposed in this first installment in this series have demonstrated that, if carefully considered, our relationships with AI products can meaningfully enhance our experiences by beginning to resemble our favorite relationships with each other, our pets, and our families.</p><p>The next installment in this series looks at how to design the methods that AI products use to get to know us better and the ways they become better at understanding and even predicting our needs. Check it out <a href="https://medium.com/@sethdanielsnyder/artificial-love-design-for-emotional-intelligence-9fa2966eb73b"><strong>here</strong></a>.</p><p>I’d like to thank my many thought partners on this project: first and foremost, my mom and Professor of Psychology at Emory University Dr.Sherryl Goodman, and my fellow <a href="https://www.frogdesign.com/">frogs</a> <a href="https://medium.com/u/788e9f70a159">Andrew Haskin</a>, <a href="https://medium.com/u/1e72b0be262b">Eve Weinberg</a>, <a href="https://medium.com/u/ca1dba03123">FabioCarnevaleMaffè</a>, Jonathan Grossman, Ara Acle, Eli Myers, Christine Todorovich, Nikki Horn, Lia Hanson, and Mayaan Albert. Thank you all for your patient editing, research support, inspiration, and advice.</p><h3>Sources and Readings:</h3><p>If you’ve read this far and still want to learn more about Artificial Intelligence and robotics I strongly recommend these sources:</p><ul><li><a href="https://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html">The AI Revolution: The Road to Superintelligence</a></li><li><a href="https://medium.com/@odannyboy/personality-not-specs-designing-social-robots-743f10bded91">Personality, Not Specs: Designing Social Robots</a></li><li><a href="http://www.npr.org/sections/alltechconsidered/2017/03/10/519002884/it-has-to-have-a-soul-how-chatbots-get-their-personalities">&#39;It Has To Have A Soul&#39;: How Chatbots Get Their Personalities</a></li><li><a href="https://blog.prototypr.io/a-guide-to-developing-bot-personalities-c6eba213d77b">A guide to developing bot personalities.</a></li><li><a href="https://medium.com/@brianroemmele/why-didnt-google-choose-a-human-like-name-for-google-home-like-amazon-did-for-echo-naming-it-alexa-87872bb220b2">Siri, Alexa, Viv, Google?: Why Google needs to give a name (and personality) to Google Home</a></li><li><a href="https://qz.com/se/machines-with-brains/1019711/what-makes-a-robot-creepy-versus-friendly-companies-are-hiring-pixar-engineers-to-figure-it-out/">Robotics companies are hiring Pixar engineers to make their robots friendlier</a></li><li><a href="http://www.ynharari.com/book/homo-deus/">Homo Deus - Yuval Noah Harari</a></li><li><a href="https://www.theatlantic.com/technology/archive/2016/03/what-is-a-human/473166/">What Is a Robot?</a></li></ul><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fupscri.be%2F8f5f8b%3Fas_embed%3Dtrue&amp;dntp=1&amp;url=https%3A%2F%2Fupscri.be%2F8f5f8b&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=upscri" width="800" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href">https://medium.com/media/c43026df6fee7cdb1aab8aaf916125ea/href</a></iframe><figure><a href="https://becominghuman.ai/artificial-intelligence-communities-c305f28e674c"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*2f7OqE2AJK1KSrhkmD9ZMw.png" /></a></figure><figure><a href="https://upscri.be/8f5f8b"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*v-PpfkSWHbvlWWamSVHHWg.png" /></a></figure><figure><a href="https://becominghuman.ai/write-for-us-48270209de63"><img alt="" src="https://cdn-images-1.medium.com/max/255/1*Wt2auqISiEAOZxJ-I7brDQ.png" /></a></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5b21c11242c8" width="1" height="1" alt=""><hr><p><a href="https://becominghuman.ai/artificial-love-designing-for-emotional-intelligence-5b21c11242c8">Artificial Love: Design for Emotional Intelligence</a> was originally published in <a href="https://becominghuman.ai/">Becoming Human: Artificial Intelligence Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>