<?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 Johannes Schleith on Medium]]></title>
        <description><![CDATA[Stories by Johannes Schleith on Medium]]></description>
        <link>https://medium.com/@jshlth?source=rss-9cd1af07b838------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Z3TisImCeM8iIGN1Z4hcYA@2x.jpeg</url>
            <title>Stories by Johannes Schleith on Medium</title>
            <link>https://medium.com/@jshlth?source=rss-9cd1af07b838------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 02:35:16 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jshlth/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[Thinking Before Prompting]]></title>
            <link>https://medium.com/@jshlth/thinking-before-prompting-6433b4fde1ab?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/6433b4fde1ab</guid>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[cognition]]></category>
            <category><![CDATA[writing-prompts]]></category>
            <category><![CDATA[conceptual-design]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Wed, 20 May 2026 17:36:38 GMT</pubDate>
            <atom:updated>2026-05-20T19:04:12.324Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/847/1*2Hsu0jAjCNDmCU1yl1dCeQ@2x.jpeg" /></figure><h3>Think Before You Prompt</h3><p>You open a new chat. You type whatever is on your mind. You hit enter. You read the output. You paste it somewhere.</p><p><strong>Sound familiar?</strong></p><p>Here’s an uncomfortable question: in that workflow — <em>who is actually doing the thinking?</em></p><h3>Where Does Thinking Happen?</h3><p>In the 1990s, cognitive anthropologist Edwin Hutchins studied how navigation crews operated US Navy ships. His conclusion was radical: <em>the mind is in the world</em>. Thinking doesn’t only happen inside individual heads — it is distributed across people, artifacts, tools, and the environment. <a href="https://learning-theories.com/distributed-cognition-dcog.html">[1]</a></p><p>This is called <strong>distributed cognition</strong>. A cockpit, a spreadsheet, a sticky note — all of these are places where thinking <em>actually happens</em>. We’ve always offloaded parts of our cognition onto the world around us.</p><p>Fine. So far, so normal.</p><p>But here’s what’s interesting. Back in 2011, Betsy Sparrow at Columbia University documented something called the <strong>Google Effect</strong>: when we expect to find information easily online, we simply don’t bother remembering it. We encode <em>where to find things</em> — not the things themselves. <a href="https://news.columbia.edu/news/study-finds-memory-works-differently-age-google">[2]</a></p><p>We’ve been training ourselves to reach outward, not inward.</p><h3>Writing a Story vs. Watching a Movie</h3><p>Let’s try a different framing. Think about what happens cognitively in each of these three situations:</p><p><strong>Writing a story</strong> — you imagine every scene, every character, every detail. The thinking is entirely yours.</p><p><strong>Reading a book</strong> — you’re prompted by the text, but the imagining still happens in your head.</p><p><strong>Watching a movie</strong> — someone else has done the imagining. You just decide what to watch.</p><p>The question for how we work with AI is: <em>should your creation process feel more like writing a story — or watching a movie?</em></p><p>When we prompt carelessly, we’re closer to the movie end of the spectrum. When we craft a prompt with intention, we’re back to something that looks like writing.</p><h3>Trial and Error vs. Prompt Crafting</h3><p>We’ve seen two distinct patterns in how people actually use LLM-powered tools.</p><p>In a <strong>‘Trial and Error’ </strong>approach users fires off short, vague prompts. See what comes back. React. Try again. Use LLMs to iterate over the prompt. Iterate toward something acceptable. It’s fast, it’s frictionless, and the output — more often than not — feels a little like it could have come from anyone.</p><p>In a <strong>‘Prompt Crafting’ </strong>approach, users spend a significant amount of time thinking <em>before</em> they type a single word. They draft bullet points. They think through structure and edge cases. They write and rewrite the prompt itself. By the time they submit, much of the real thinking has already been done — the LLM is just doing the polishing.</p><p>One could argue, that the difference in <em>ownership</em>. ‘Trial and Error’ hands over authorship to the tool, one iteration at a time. ‘Crafting’ maintains some conceptual control.</p><h3>Conceptual vs. Detailed Design</h3><p>In UX, we used to separate <strong>low-fidelity conceptual design</strong> from <strong>high-fidelity detailed design</strong>. The whole point of a wireframe — a literal boxes-and-lines sketch — was to force a conversation about <em>structure and intent</em>, before anyone invests time in polish.</p><p>Interestingly, rough prototypes consistently generate better, more honest feedback than polished ones. Stakeholders comment on substance when they can’t be distracted by surface — and when they need to ‘imagine’ more with a critical mind. <a href="https://www.interaction-design.org/literature/topics/low-fidelity-prototyping">[3]</a></p><p>Maybe the same principle applies to prompts. Before we share the polished AI output with our team, should we be sharing the bullet-point outline? Iterating on the <em>intent</em> first — and leaving the formatting to later?</p><p><em>Share the thinking. Not just the result.</em></p><h3>Surrender vs. Delegation</h3><p>Shaw and Nave introduce a useful distinction <a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646">[4]</a>. There’s a continuum, they argue, that starts with <strong>cognitive offloading</strong> — using a tool strategically, the way you use a calculator. You’re still in charge of judging whether the result makes sense.</p><p>Then there’s <strong>cognitive surrender</strong> — the near-total abdication of independent reasoning. You adopt the output, no questions asked, no effort invested. The dial moves from thinking <em>with</em> AI to letting AI think <em>instead of</em> you.</p><p><em>The alarming part? It feels like progress.</em> — Psychology Today, Feb 2026</p><p>The study found that participants followed AI outputs 80% of the time — even when the AI was <em>deliberately wrong</em>. The signal gets drowned out when we stop applying our own judgment.</p><p><em>Delegation is fine. Surrender is not.</em></p><h3>Think, Refine, Prompt</h3><p>So what does this mean when using AI for product development and design?</p><p>When prototyping with AI tools like v0, Replit, or Figma Make — you could fire off one tiny prompt at a time and let the machine figure it out. Or you could spend time imagining the user journey first, sketching out the key interactions in bullet points, and <em>then</em> translate that thinking into a prompt. The first approach is faster to start. The second approach puts the conceptual design work where it belongs — in your head.</p><p>Same goes for a product requirements document. You could hand the AI a vague brief. Or you could go through several iterations of your own thinking — write a messy bulleted list, interrogate it, restructure it — and <em>then</em> ask the LLM to help you shape it.</p><p>I’d argue, the bullet point prompt, the imagination before putting pen to paper is the actual work.</p><p>Maybe we should get more comfortable sharing rough notes and unpolished thinking — and treat polished AI output as a final presentation layer, not the process itself.</p><p><em>Do you tend to trial and error your way through prompts? Or do you sit with the thinking first?</em></p><p><strong>Read more:</strong></p><ol><li><a href="https://learning-theories.com/distributed-cognition-dcog.html">Distributed Cognition — Edwin Hutchins, Cognition in the Wild (1995)</a></li><li><a href="https://news.columbia.edu/news/study-finds-memory-works-differently-age-google">Google Effects on Memory — Sparrow et al., Science (2011)</a></li><li><a href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646">Thinking — Fast, Slow, and Artificial — Shaw &amp; Nave, Wharton (2026)</a></li><li><a href="https://www.psychologytoday.com/us/blog/harnessing-hybrid-intelligence/202602/are-we-cruising-toward-cognitive-capitulation">Are We Cruising Toward Cognitive Capitulation? — Psychology Today (2026)</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6433b4fde1ab" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Don’t design for AI]]></title>
            <link>https://uxdesign.cc/dont-design-for-ai-part-1-01b2d5f936ae?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/01b2d5f936ae</guid>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Mon, 19 Feb 2024 18:48:33 GMT</pubDate>
            <atom:updated>2024-02-29T19:43:41.143Z</atom:updated>
            <content:encoded><![CDATA[<h4>Design for problems worth solving with AI, distributed cognition, cognitive tasks and <em>human centered</em> intelligent systems</h4><p>How often have we been approached by a client to “do the AI” in a project… with the expectation that some (generative) AI might “magically” solve all known, unknown, and mostly undefined problems?</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2FQjdV5Z6ZTUmloHIBWX%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExZW41bXpqYmcxbHluenRtdDRtNHVmNG1leWxwMGd1MWtmanViaHVtcCZlcD12MV9pbnRlcm5hbF9naWZfYnlfaWQmY3Q9Zw%2FQjdV5Z6ZTUmloHIBWX%2Fgiphy.mp4&amp;image=https%3A%2F%2Fmedia3.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExYmsxbHcyemJwNnp0ZGU2aGNoMjBoMmRnaGhjM282eXE3bzYxMDd3dCZlcD12MV9naWZzX2dpZklkJmN0PWc%2FQjdV5Z6ZTUmloHIBWX%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="246" frameborder="0" scrolling="no"><a href="https://medium.com/media/fc51ccceb9175cc5a68346b005c3282f/href">https://medium.com/media/fc51ccceb9175cc5a68346b005c3282f/href</a></iframe><p>In this opinionated article, based on a talk at <a href="https://techcircus.io/en/events/ux-live-london">UX Live 2023</a> by <a href="https://medium.com/u/22c36a2e9cc5">Jacqueline Martinez</a> and <a href="https://medium.com/u/9cd1af07b838">Johannes Schleith</a>, we propose tweaking design thinking methods for opportunity detection.</p><blockquote>When you have a Large Language Model, everything looks like an opportunity for a chat-based assistant …</blockquote><h3>Design for problems worth solving!</h3><p>Let’s go back to the <a href="https://www.designcouncil.org.uk/our-resources/framework-for-innovation/">design basics</a>. Rather than jumping to solutions, a well-structured problem-led approach aims to investigate and clearly define problems (Problem Discovery), prior to experimentation with various possible solutions to those problems (Solution Exploration).</p><figure><img alt="problem discovery phase followed by solution exploration phase" src="https://cdn-images-1.medium.com/max/1024/1*QLCxaSj-N3RueVY0k3JSWA.png" /></figure><p>Not every problem needs to be solved with AI. Not every AI opportunity needs to be solved with a <a href="https://en.wikipedia.org/wiki/Large_language_model">Large Language Model (LLM)</a>. Focusing on the problem first allows to explore alternative solutions, benchmark and evaluate them and pick the most suitable solution.</p><p>With that in mind, we propose to structure <strong>AI opportunity detection</strong> with frameworks such as <a href="https://en.wikipedia.org/wiki/Distributed_cognition"><strong>distributed cognition</strong></a> and <a href="https://dl.acm.org/doi/fullHtml/10.1145/3527927.3532808"><strong>cognitive strategy prompts</strong></a><strong>, </strong>and conduct <strong>user research</strong> and <strong>data landscaping</strong> in parallel. We’ll define these terms below…</p><figure><img alt="AI opportunities are found at the intersection of problem space and information space. Problems that involved cognitive tasks are “worth solving with AI”. Problems for which sufficient data is available are “feasible to be solved with AI”" src="https://cdn-images-1.medium.com/max/1024/1*1Fr7fPf0BXKpGFwf29U5uQ.png" /><figcaption>Illustration of the sweet spot of problems worth solving and problems that are feasible to be solved with AI</figcaption></figure><h3>Design for Distributed Cognition!</h3><p>In a nutshell, <a href="https://en.wikipedia.org/wiki/Distributed_cognition">Distributed Cognition</a> is an established <a href="https://www.interaction-design.org/literature/topics/human-computer-interaction">Human Computer Interaction</a> (HCI) framework, which allows teams to take on a holistic perspective beyond technology and software. One of the key ideas is that “cognition” is not limited to human brains only, but an emergent property that is dispersed across people, artefacts and technologies.</p><figure><img alt="Illustration of distributed cognition. A client ask a lawyer to review “repair obligations” in a “commercial lease”. The lawyer collaborates with a paralegal via a matter management system, in order to identify and evaluate the repair clause. legal advice is provided back to the client. Different actors (client, lawyer, matter management system, paralegal) need to collaborate and “think together” in order to accomplish the task." src="https://cdn-images-1.medium.com/max/1024/1*e7-pw2pzF89marmBuMSaMg.png" /><figcaption>Illustration of information flows between client, laywer and matter management system</figcaption></figure><p>In our illustration, client, lawyer, matter management system and paralegal are collaborating, in order to satisfy the initial client request. Each actor uses their knowledge (business, legal, data) in order to manipulate the available data, provide cognitive services (review, summary, advice) and ultimately generate legal advice.</p><p>It can be quite tricky to fully describe a complete picture of such collaborative systems. However, user research and systems thinking methods can helps us understand and map processes, goals and describe in detail how actors collaborate and “think together”.</p><p>Once we understand the system, we can detect AI opportunities!</p><h3>Design for Cognitive Tasks</h3><p>In order to map the process of distributed cognition, we capture the expected input, desired output, and core <a href="https://jobs-to-be-done.com/">jobs-to-be-done (JTBD)</a>.</p><figure><img alt="Work of individual actors (e.g. paralegal) can be mapped as an input, process, output flow. In addition we consider “expected input data” (e.g. commercial lease and the question for the repair obligation) and “desired output data” (e.g. evaluation of the repair clause). It can further be broken down into jobs-to-be-done and jobs that involve specific “cognitive tasks”, such as “extraction”, “comparison” or “review” of information (e.g. extraction of the repair clause)" src="https://cdn-images-1.medium.com/max/1024/1*A0fjC2Y7QxM80SDM4lZWkw.png" /><figcaption>Illustration, from process mapping to cognitive tasks</figcaption></figure><p>In a second round, we try to reframe JTBD into “cognitive tasks” through a number of <a href="https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9">cognitive strategy prompts</a>. We might ask what kind of “information handling” activites are involved in the process, whether jobs are related to search, lookup, review, contrasting or relation of information?</p><figure><img alt="Illustration of 10 cognitive strategy prompts: get an overview on a domain, looking for unknowns, lookup known information, re-find information, differentiate or relate information, connect data to (domain) expertise, monitor a domain, extract information, create information and summarisation" src="https://cdn-images-1.medium.com/max/1024/1*NcyPTpV7wn_qqhZsXaiSAQ.png" /><figcaption>Illustration of Cognitive Strategy Prompts</figcaption></figure><p>We found that application of the prompts can significantly increase the number of ideas that are AI relevant, because they nudge ideation towards activities that can actually be supported with AI methods.</p><p>Now that we understand the process, information and cognitive tasks, its about time to look into the data that can be used for experimentation</p><h3>Design based on the Data Landscape</h3><p>We can (user) research and design all we want, if we ignore the constraints — or opportunities — of underlying data we will not create successful AI powered innovation. After all, AI is only as good as the data you train it on.</p><figure><img alt="Each domain has an information space that consists of known data sets (e.g. commercial lease and repair clause), domain knowledge that is not represented in digital data (e.g. definition of a repair clause) and other not yet known data sets (e.g. body of a collection of commercial leases)" src="https://cdn-images-1.medium.com/max/1024/1*sTIfgKb2za9bP9hRibbyuQ.png" /><figcaption>Illustration of the information space and data landscape</figcaption></figure><p>Inspired by the <a href="https://medium.com/disruptive-design/quick-guide-to-the-disruptive-design-method-25a7b2b56e5f">Disruptive Design Method</a> we aim to get a holistic, system level view on the landscape of <strong>all information involved in our use case. </strong>What data are available? How can they be accessed? What specific data formats are relevant to our use case?</p><p>We might explore beyond whats in front of our eyes and think about what other data sources that we haven’t considered yet might be relevant?</p><p>Often “information” that is highly relevant to a use case is not actually present in digital data — but resides in individuals heads, in the form of domain knowledge.</p><p>For example, in legal tech innovation, domain experts are highly trained individuals with rich taxonomies, precedent cases and matters deeply memorized in their human brains. Its crucial to be aware of such “data” and consider how to enable end users to bring in their very own expert view.</p><p>We are now ready to transform our AI opportunities into hypotheses and experiments.</p><h3>Why experiment?</h3><p>How much resource is wasted by locking-in to one specific technology or concept too early, while missing to evaluate product market fit and ignoring alternatives?</p><blockquote>Even if you run as fast as you can, if you are going in the wrong direction, you’ll fail!</blockquote><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fgiphy.com%2Fembed%2F4QdAF8Mby9vW0%2Ftwitter%2Fiframe&amp;display_name=Giphy&amp;url=https%3A%2F%2Fmedia.giphy.com%2Fmedia%2F4QdAF8Mby9vW0%2Fgiphy-downsized.gif&amp;image=https%3A%2F%2Fmedia4.giphy.com%2Fmedia%2Fv1.Y2lkPTc5MGI3NjExNHg0bjcyYmdjOTBkamlxd2JmdGIzajdpNzNycHNpa3hzYXgxbWo0OSZlcD12MV9naWZzX2dpZklkJmN0PWc%2F4QdAF8Mby9vW0%2Fgiphy.gif&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=giphy" width="435" height="283" frameborder="0" scrolling="no"><a href="https://medium.com/media/55b1c89ab46c941fe4cae4475f09f44f/href">https://medium.com/media/55b1c89ab46c941fe4cae4475f09f44f/href</a></iframe><h3>Reframe Data Science Problems</h3><p>First, we want to <strong>reframe </strong>what we know about <a href="https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9"><strong>cognitive tasks</strong></a> as data science problems. Data science is what makes AI work and tick. So let’s map user needs to specific data science problems. Data science</p><p>The better we can define expected input data (e.g. real written reports that need to be summarised), its manipulation and desired output (e.g. some real report summaries), the easier it is for data scientists to work through some experiments.</p><figure><img alt="Diagram with 3 examples how cognitive strategy prompts map to different data science methods. for example the desire to “get an overview on information” can be supported by data science methods such as knowledge graphs, clustering or topic modelling. Whereas lookup of information can be supported by search methods, infromation extraction or named entity recognition." src="https://cdn-images-1.medium.com/max/1024/1*Z5er_lN7_ixZzHd2AY4rQQ.png" /><figcaption>Illustration of some examples how users cognitive tasks might be reframed as data science problems</figcaption></figure><p>For example, imagine that a legal researcher wants to <em>“get an overview on information”</em>, related to a legal case, their client’s industry or a new practice area they are working on — we could assist this task with various data science methods — depending on what our data allows for — e.g. <em>knowledge graphs, clustering techniques, anomaly detection, topic modelling</em> and so on</p><p>While our <a href="https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9">paper</a> explored a mapping between prompts and data science methods, such mappings will always be incomplete due to the ever evolving AI methods and their surprising emerging qualities. It is crucial to involve data scientists early in the ideation process — precisely define the problem for them and work through various alternative solutions, so that we can evaluate different ideas and hypotheses and select the best fitting solution to our problem.</p><blockquote>“AI is not one thing […] it’s only AI until you understand it — then it’s just software”</blockquote><h3>Design for Human Centered AI</h3><p>Human centred design for software products is table stakes. Many successful products have been built with a strong focus on user experience and usability. The same should be true for AI powered products.</p><p><a href="https://global.oup.com/academic/product/human-centered-ai-9780192845290">Human Centered AI</a> (HCAI) is an emerging and quickly evolving field. that prioritizes human aspects like human control, human-in-the-loop and feedback loops, trust, error recovery, explainability, and many more.</p><figure><img alt="Overview on some human centered AI themes, such as human control, error recovery, expectations, privacy, explainability, feedback loops, anthromorphism, accessibility, novelty. Each of which might be considered during the design and definition of an AI powered solution." src="https://cdn-images-1.medium.com/max/1024/1*fRXH01xrn8Iyuw__GDcN6Q.png" /></figure><p>When exploring alternative solutions to a well-defined problem, during a conceptual design phase, it is worth working through some of these aspects and consider how to optimize the design for them.</p><p>For starters, this can be as simple as taking a conceptual prototype or wireframes into a working session and thinking through one theme at a time, and how to optimize for such a theme.</p><p>Let’s look at two examples: Explainability and Human-in-the-loop.</p><h4>HCAI Example: Design for Explainability</h4><p>A great example is the 2021 <a href="https://dl.acm.org/doi/10.1145/3411763.3443441">paper</a> by Norkute et al. on explainability for a text summarisation solution. The team experimented with methods to summarise multiple pages of legal cases into one (or a few) sentences. The goal to optimize the system for <em>explainability</em> made the team of researchers consider and evaluate alternative approaches to explain how the system had generated each summary.</p><p>While one explanation of the generated summary used an off-the-shelf explainability method that highlighted the section that was most important source for the summary, another method put more effort into calculating attention vectors and highlighting a heatmap of words that were most relevant for the summarisation. Both approaches could be evaluated and A/B tested.</p><p>Note that the best explanations do <em>not</em> go into unnecessary detail, but provide some <em>easy-to-grasp layer of communication</em> that gives the user an <em>idea</em> of how the system might behave. On the flip side, over-explaining with too much (technical) detail can easily overwhelm non-technical end users, and might be confusing rather than helpful.</p><figure><img alt="Illustration of two approaches how to explain a summarisation algorithm. “Source highlighting” surfaces only the most relevant sentence in the source text. “Attention highlighting” shows a heatmap of words in the source text that have impacted the summary. The latter performed better during user tests." src="https://cdn-images-1.medium.com/max/1024/1*RzYY_73tgHUE0tuGPfsWZw.png" /><figcaption>Illustration, explain summarization through highlighting with attention vectors (left) or source highlighting (right). Attention vector highlighting was found to be more useful to end users.</figcaption></figure><h4>HCAI Example: Design for Human-in-the-Loop</h4><p>In his book <a href="https://global.oup.com/academic/product/human-centered-ai-9780192845290?cc=us&amp;lang=en&amp;">Human Centered AI</a>, Ben Shneiderman showed that increased levels of autonomy by the system, do not necessarily have to mean decreased levels of control by a human.</p><p>Alongside optimizing for explainability, it is worth to consider ideation on concepts for feedback loops. Often it is crucial to keep human end users informed and up-to-date about the current behaviour of the system — or provide detailed audit trails and logs to monitor decisions by the system in hindsight.</p><p>Always keep in mind that, while seemingly <em>intelligent</em>, AI systems are nothing other than statistical machines that easily mistake meaning of words or symbols that violate common sense.</p><figure><img alt="Wireframes of a user interface that allows the user to provide feedback on a classification system. For classified items, the user can provide detail on what went wrong and whether the suggested class was correct or not." src="https://cdn-images-1.medium.com/max/1024/1*nNQdlkHH-jFxdZ6w-eGSrA.png" /></figure><h3>Design for hypothesis testing</h3><p>Each idea, how we might reframe a cognitive task as a data science problem, is essentially a hypothesis we can test!</p><p>Depending on your team size, you can be more or less formal about defining hypothesis, expected input data, desired output data, core jobs to be done to support, desired outcomes for the end user etc. etc. Data scientists can then experiment with various data science methods and AI methods — as long as they are provided with relevant (training) data.</p><figure><img alt="Example of a test card that asks participants to reflect and define an experiment. It encourages participants to describe their hypothesis, how to test it, what metric to measure and what criteria to use in order to know whether the experiment was successful or not." src="https://cdn-images-1.medium.com/max/478/0*DlTYXnqF73HAf_xe.png" /></figure><p>For example, <a href="https://www.strategyzer.com/library/validate-your-ideas-with-the-test-card">Strategyzer’s test card</a> is an easy to grasp artefact that can be used to reflect on and discuss hypotheses and experiment to evaluate them in a workshop session.</p><h3>Prototyping for Dynamic Content</h3><p>In addition to technical proof of concepts, we might also want to test hypotheses with domain experts and end users through rapid prototyping and concept testing.</p><p>While Machine Learning is a <a href="https://dl.acm.org/doi/10.1145/3025453.3025739">“new design material”</a><em>, </em>prototyping with data and dynamic content brings its very own challenges. Current design tools (e.g. Figma, Framer, Sketch, Axure etc.) are fantastic to prototype and test user journeys — as long as the content within these journeys remains <em>static</em>. In contrast, often the very thing we want to evaluate is the quality of the machine generated, <a href="https://nips.cc/virtual/2021/48930">dynamic content</a>.</p><blockquote>UX professionals need to play with code and get their hands dirty!</blockquote><p>It’s worth working closely with data scientists to test their early results, either straight out of the computer terminal, <a href="https://jupyter.org/">Jupyter notebook</a>, <a href="https://plotly.com/">plotly data visualisation</a> or wrap a simple <a href="https://streamlit.io/">Streamlit</a> application around their python functions.</p><figure><img alt="Screenshots of two prototyping environments, Streamlit a python library and a jupyter notebook that uses plotly to display some data visualisation." src="https://cdn-images-1.medium.com/max/1024/1*UTahzSyLBceJ_JHL-JG3TQ.png" /><figcaption>Illustration of streamlit, jupyter and plolty</figcaption></figure><p>User testing straight out of such prototypes allows to evaluate the AI algorithms’ output and can be combined with UX prototypes that <em>tell the story </em>of the product around the content. Early evaluation often also inspires design changes to find a conceptual solution that better fits the dynamically generated content.</p><p>Where this is not possible, good old HCI techniques like <a href="https://www.nngroup.com/articles/wizard-of-oz/">Wizard of Oz</a>, <a href="https://www.alexandercowan.com/creating-a-lean-startup-style-assumption-set/">Concierge testing</a> or similar, can go a long way to provide useful end user evaluation in early stages of a project.</p><p>Aspects to test for might encompass perceived trust, perceived control and usability, perceived capability and intelligence, and address the ultimate question of <em>how </em>a new AI-powered product might fit into the end users day-to-day.</p><p>To wrap things up. We recommend to tweak design process and methods in the following way…</p><ul><li>a good understanding of how <strong>distributed cognition</strong> happens throughout the system or use case under investigation.</li><li>a clear focus on <strong>“cognitive tasks”</strong> and <strong>“information handling” </strong>in specific processes related to core <strong>jobs-to-be-done</strong></li><li>a <strong>data inventory</strong> and understanding the constraints of <strong>available and accessible data sets</strong></li><li>detect <strong>AI opportunities</strong> through a focus on<strong> distributed cognition</strong> and <strong>cognitive tasks</strong>, in conjunction with an open minded <strong>data inventory</strong></li><li>reframe user needs into<strong> data science problems</strong></li><li>design for <strong>Human-centered AI</strong> and prototype with <strong>dynamic data</strong> to <strong>test your hypotheses</strong></li></ul><p>We hope this mini series was useful to you. Let us know in a comment, whether you agree, challenge any of the points or want to add any thought!</p><p><strong>Read more:</strong></p><ul><li><a href="https://www.designcouncil.org.uk/our-resources/framework-for-innovation/">Design Council: Double Diamond</a></li><li><a href="https://en.wikipedia.org/wiki/Distributed_cognition">Distributed cognition</a></li><li><a href="https://jobs-to-be-done.com/">Jobs-to-be-done (JTBD)</a></li><li><a href="https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9">Cognitive Strategy Prompts for Human Centered AI Opportunity Detection [Medium]</a></li><li><a href="https://dl.acm.org/doi/fullHtml/10.1145/3527927.3532808">Cognitive Strategy Prompts: Creativity Triggers for Human Centered AI Opportunity Detection [ACM]</a></li><li><a href="https://medium.com/disruptive-design/quick-guide-to-the-disruptive-design-method-25a7b2b56e5f">Disruptive Design Method</a></li><li><a href="https://dl.acm.org/doi/10.1145/3411763.3443441">Towards Explainable AI: Assessing the Usefulness and Impact of Added Explainability Features in Legal Document Summarization</a></li><li><a href="https://hyperight.com/explainability-for-text-summarization-of-legal-documents-nina-hristozova-milda-norkute-thomson-reuters/">https://hyperight.com/explainability-for-text-summarization-of-legal-documents-nina-hristozova-milda-norkute-thomson-reuters/</a></li><li><a href="https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9">Cognitive Strategy Prompts for Human Centered Opportunity Detection</a></li><li><a href="https://global.oup.com/academic/product/human-centered-ai-9780192845290">Ben Shneiderman — Human Centered AI</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=01b2d5f936ae" width="1" height="1" alt=""><hr><p><a href="https://uxdesign.cc/dont-design-for-ai-part-1-01b2d5f936ae">Don’t design for AI</a> was originally published in <a href="https://uxdesign.cc/">UX Collective</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cognitive Strategy Prompts for Human Centered AI Opportunity Detection]]></title>
            <link>https://medium.com/@jshlth/cognitive-strategy-prompts-for-human-centered-ai-opportunity-detection-c052a9af74c9?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/c052a9af74c9</guid>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[human-centered-ai]]></category>
            <category><![CDATA[user-experience-research]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[innovation]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Sun, 11 Dec 2022 15:56:48 GMT</pubDate>
            <atom:updated>2022-12-11T15:59:27.760Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aNjaMGCPcONzoDVTT_WGvg.png" /></figure><p><em>This article is based on the publication </em><a href="https://dl.acm.org/doi/abs/10.1145/3527927.3532808"><em>“Cognitive Strategy Prompts” (Creativity &amp; Cognition 2022, ACM)</em></a><em>, co-written with </em><a href="https://medium.com/u/9518f01e9857"><em>Daniella Tsar</em></a><em>, </em><a href="https://medium.com/u/e3beeb42cada"><em>Milda Norkute</em></a><em> and Mary Mikhail.</em></p><p>When it comes to innovating a service or business, it seems like “AI” is the thing to do in 2023. However, instead of further defining the problem, we find our stakeholders play <strong>AI buzzword bingo</strong> by throwing the latest and greatest trending <strong>“AI terms”</strong> into the conversation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/215/1*v2IaZ4i_ziec5VzeFDiiPg.png" /><figcaption>AI / ML buzzword bingo</figcaption></figure><p>Even worse, our client might expect us to <strong>“just apply some AI Magic”</strong> to solve any pain point and bottleneck that might occur in their business.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/406/1*mj6GEB5uE-0bHkjlZQyx8Q.png" /><figcaption>Forget about the problem, apply some AI Magic…</figcaption></figure><p>So, how might we identify precise, actionable and human-centered problems that are worth solving with AI methods? How can we adapt user research activities and design process to facilitate opportunity detection for AI powered innovation?</p><h3>Cognitive Strategies</h3><p>There is plenty of terminology to describe AI systems and features.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/330/1*024l9aeiWmTERkyFoiiQKw.png" /><figcaption>Illustration AI/ML terminology</figcaption></figure><p>However, we have much less words to describe and decipher the processes and kinds of knowledge work we want to assist.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/382/1*mb444Ls3Y_DRFZLeMW4H0w.png" /><figcaption>Lack of terminology to describe knowledge work</figcaption></figure><p>Knowledge workers <strong><em>Search</em></strong> for, <strong><em>Investigate</em></strong>, and<strong> <em>Make decisions</em></strong> based on information on a daily basis. Based on the categorisation of information retrieval needs and cognitive processes, we structure information seeking, obtaining and handling of information into 5 different categories and prompts: <strong><em>Learn, Lookup, Relate, Monitor &amp; Extract, </em></strong>and<strong><em> Create.</em></strong></p><p>Each category has its unique cognitive strategies and might pose different challenges for the application of data science methods. For example to <strong><em>differentiate, contrast </em></strong>or<em> </em><strong><em>relate </em></strong>data points when <strong><em>investigating</em></strong> information (<a href="https://dl.acm.org/doi/abs/10.1145/3527927.3532808">see paper for full list of prompts</a>).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aNjaMGCPcONzoDVTT_WGvg.png" /><figcaption>Cognitive Strategy Prompts</figcaption></figure><h3>Application</h3><p>After all, AI is defined as a methodology that <em>‘</em><a href="https://www.ibm.com/uk-en/cloud/learn/what-is-artificial-intelligence"><em>aims to mimic the problem-solving and decision-making capabilities of the mind</em></a><em>’.</em> The proposed methodology of cognitive strategy prompts gives us a clear and actionable linguistic platform to link the human cognitive processes to the corresponding AI methodologies.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/1*Khzw0AS5hsijx_SnWkce5A.png" /><figcaption>Application of Cognitive Strategy prompts to refine user needs or reframe ideas into data science problems</figcaption></figure><p>Prompts might support initial discovery and user research, serve as prompts during workshops, or facilitate user needs mapping and reframing of data science problems.</p><h3>Cognitive Strategy Prompts for user research</h3><p>The prompts aim to support user research, e.g. creation of moderation guides, probing during interviews, as well as analysis such as process mapping, jobs-to-be-done or task analysis.</p><p>Let’s look at an example.</p><blockquote>Let’s imagine that we want to research how travellers plan and book holidays. We might use the prompts to investigate whether travellers are “trying to get an overview” on travel options, or whether they “search” for specific holiday destinations.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/1*PoeWbsNFxNnHgNIQG335Ng.png" /><figcaption>Illustration of the application of a cognitive strategy prompt</figcaption></figure><p>We can also apply prompts in participatory co-creation workshops. During our experimental sessions, we conducted a jobs writing exercise, in which we asked participants to write jobs-to-be-done for a given scenario, e.g. activities that are involved in the “travel planning”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/1*Ug6HAvPe3Vl-jwXu5eWYHg.png" /><figcaption>Example workshop agenda</figcaption></figure><p>After an initial understanding of process and tasks was established, we introduced the prompts. Participants were asked to think again, which “cognitive strategies” they might apply at which stage and refine the jobs accordingly. The refined jobs statements were used for further problem definition and ideation.</p><h3>Reframe Prompts into Data Science Problems</h3><p><em>“When you have a hammer everything looks like a nail”. </em>It can be challenging to translate conceptual ideas into data science problems. Often, stakeholders might <strong>“lock into known AI buzzwords” </strong>(e.g. “I heard about deep learning, can we solve XYZ with deep learning?”), scientists might be biased by methods they recently used on other projects.</p><p>In a survey we asked data scientists to associate each prompt to a number of data science and AI methods.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/420/1*kKP5sL3gZI_lkUeF6pjJBQ.jpeg" /><figcaption>Association of Data Science methods for each Cognitive Strategy prompt</figcaption></figure><p>The resulting mapping aims to provide options and help translate between scientists and non-technical stakeholders.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/1*V-5WDjDJN8rmH5-1Gc7eEw.png" /><figcaption>Illustration mapping Cognitive Strategy prompts to data science methods</figcaption></figure><p>When reviewing job stories, a data scientist might use the mapping to generate alternative solutions that might support the user need and underlying cognitive strategy. In reverse, they might also think through what cognitive strategies a data science method might enable.</p><p>Another example:</p><blockquote>When a news reporter is trying “to get an overview” on current events and news updates, she might well benefit from a system that applies methods such as clustering, anomaly detection or topic modelling. Ideally, this might mimic and support her approach to identify themes and relationships within news events. Different experiments with available data sets can explore which direction might be most promising for the use case.</blockquote><h3>Considerations</h3><p>Initial experiments have been overwhelmingly positive. We are aiming to expand on the cognitive strategy prompts considering the various nuances within industries.</p><p>The cards have been developed in the context of knowledge work and research for B2B services. Cognitive strategies might be different for different industries and domains, as well as for consumer products and services.</p><p>We experienced that its beneficial to work with a (very) limited set of prompts. Even the set of 10 cards proved to be challenging during some of our workshops.</p><p><em>More detail in the paper</em> <a href="https://dl.acm.org/doi/abs/10.1145/3527927.3532808"><em>“Cognitive Strategy Prompts” (Creativity &amp; Cognition 2022, ACM)</em></a><em>.</em></p><p>We are keen to improve the framework! Let us know your thoughts!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c052a9af74c9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Fluency Workshop]]></title>
            <link>https://medium.com/@jshlth/ai-fluency-workshop-5bbe1cda2b9b?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/5bbe1cda2b9b</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[workshop]]></category>
            <category><![CDATA[innovation]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Mon, 18 Oct 2021 19:01:10 GMT</pubDate>
            <atom:updated>2021-10-18T19:01:10.037Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>So you want to be an AI Innovator?</strong> Do you know where to apply AI within your domain? How can you tell what is the right approach?</p><p>Seeing groups of participants grow their ideas organically, balance and rebalance their approach was the highlight for us at the workshop we held at the <a href="https://app.swapcard.com/event/the-ai-summit-london-1/planning/UGxhbm5pbmdfNDU0ODQx">AI Summit</a> London.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JY-p2fG2CkpIgimsGDcc7w.jpeg" /><figcaption>Lean AI Design Workshop at AI Summit 2021</figcaption></figure><blockquote>If you want to stay tuned and learn more about AI Fluency, sign up, and tell us more here <a href="https://bit.ly/aifluency">https://bit.ly/aifluency</a></blockquote><h4><strong>Definition &amp; Alignment</strong></h4><p><a href="https://www.gartner.com/smarterwithgartner/3-barriers-to-ai-adoption"><strong>AI opportunity detection is poorly understood</strong></a><strong> [1].</strong> AI use case definition can be a struggle for innovation efforts. Ideation on AI challenges more often than not lead us to overly general statements, with buzzwords aplenty, and a mind locked into exciting, but narrow use of a specific technology.</p><p><strong>Great products are built by <em>interdisciplinary</em> teams.</strong> Successful teams combine subject matter experts’ domain knowledge and technical expertise in engineering and data science. Each discipline brings a wealth of invaluable knowledge to the conversation, yet each come at it from a different place.</p><p><em>‘I want actionable insights to increase customer satisfaction!’</em> exclaim the first. <em>‘Using anomaly detection would be so powerful with this data.’</em> dream the second. The first statement is so general as to be meaningless — the second locks us into a specific solution way before we have discovered the right problem to solve.</p><p>Time and time again, we have seen this phenomenon play out in settings where the difficulty of getting to a pragmatic conversation about AI is misinterpreted as ‘mismatch in technical knowledge’. It’s far from that, it’s simply the wrong level of detail about the problem at hand.</p><h4><strong>Cognitive Work Analysis</strong></h4><p>We believe that great AI opportunities are best detected by decomposing a problem into specific <a href="https://jobs-to-be-done.com/">jobs-to-be-done</a> [2] and <a href="http://faculty.washington.edu/fidelr/RayaPubs/MultiDimensionalApproach.pdf">cognitive tasks</a> [3], that lend themselves well for AI automation. By combining Design Thinking techniques with Data Science expertise we are creating a framework that facilitates a conversation between technologists and domain experts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GW5rRW4IymNNhlw1B2-PXQ.png" /><figcaption>Cognitive Work Analysis, Source: <a href="http://faculty.washington.edu/fidelr/RayaPubs/MultiDimensionalApproach.pdf">Fidel et al. — A Multidimensional Approach to the Study of Human- Information Interaction: A Case Study of Collaborative Information Retrieval</a> [3]</figcaption></figure><p>During the workshop Lean AI at the AI Summit 2021, we trialed an early version of our approach. Participants were guided through a series of exercises as a form of a framework that would provide help. First, we investigate the work environment and map stakeholders in our system …</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/474/1*NtigKY156T5tGyN831h0cA.png" /><figcaption>Illustration Stakeholder Map</figcaption></figure><p>… we then decompose a scenario into process, tasks and actionable insights per stakeholder…</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CsCFIOvKe_NA73nG7TssHQ.png" /><figcaption>Illustration Process Map</figcaption></figure><h4>Reframe Tasks as AI Problems</h4><p>Zooming into specific stages, we investigate which <strong>input data</strong> a subject matter expert uses currently? Which <strong>tasks</strong> do they carry out? What artefacts do they work with? What’s the <strong>desired output</strong> of the activity? What’s that mysterious process about? How can we more clearly define <strong>cognitive tasks</strong>, and start reframing user needs as <strong>AI problems</strong> …</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EvNr0vJtP-VpKiw1hU5FSA.png" /><figcaption>Illustration Reframing Cognitive Tasks as AI Problems</figcaption></figure><blockquote>‘It was so interesting to see, that one process or problem has so many smaller processes and components within it — and what’s important is to choose the right one’ AI Summit 2021 participant</blockquote><blockquote>‘It didn’t feel like two hours at all’ AI Summit 2021 participant</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*434AqZv0SXUL8JRTEi8G3w.png" /><figcaption>Daniella</figcaption></figure><h4>About Us</h4><p>We are two innovators, bringing in different perspectives to AI Innovation …<br><a href="https://medium.com/u/9518f01e9857">Daniella Tsar</a> (@dani_tsar) with a background in Data Science, Leading Data Science innovaviton</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/150/1*A8t_c3cmBLIFOQd9VAVFZQ.png" /><figcaption>Johannes</figcaption></figure><p><a href="https://medium.com/u/9cd1af07b838">Johannes Schleith</a> (@jshlth) with a focus on Design Thinking, User-centered Design</p><p>We are passionate about researching and experimenting with the facilitation and translation between AI Technologists and (non-technical) Subject Matter Experts.</p><h4>Stay tuned</h4><p>If you want to stay tuned and learn more about AI Fluency, sign up, and tell us more here <a href="https://bit.ly/aifluency">https://bit.ly/aifluency</a></p><h4>References</h4><ol><li>3 Barriers to AI Adoption, <a href="https://www.gartner.com/smarterwithgartner/3-barriers-to-ai-adoption">https://www.gartner.com/smarterwithgartner/3-barriers-to-ai-adoption</a></li><li>Jobs to Be Done: Theory to Practice Book by Anthony W. Ulwick, <a href="https://jobs-to-be-done.com/">https://jobs-to-be-done.com/</a></li><li>Fidel et al. — A Multidimensional Approach to the Study of Human- Information Interaction: A Case Study of Collaborative Information Retrieval, <a href="http://faculty.washington.edu/fidelr/RayaPubs/MultiDimensionalApproach.pdf">http://faculty.washington.edu/fidelr/RayaPubs/MultiDimensionalApproach.pdf</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5bbe1cda2b9b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Don’t Overlawyer]]></title>
            <link>https://medium.com/@jshlth/dont-overlawyer-1b12c77258b8?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/1b12c77258b8</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[legaltech]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[innovation]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Tue, 06 Apr 2021 11:14:20 GMT</pubDate>
            <atom:updated>2021-04-06T11:14:20.463Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*WMvVJaKvP3qJ36PVE9HpJA.jpeg" /></figure><p><em>What’s the right balance between being correct — and being able to act?</em></p><p>This article is co-written by <a href="https://medium.com/u/9518f01e9857">Daniella Tsar</a> and <a href="https://medium.com/u/9cd1af07b838">Johannes Schleith</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*Jr3c_kIhB8TTA0vnX4AmCA.png" /></figure><h3>Don’t Overlawyer</h3><p>Our team regularly holds customer workshops to support ideation and problem solving with professionals in legal, tax, compliance and accounting.</p><p>During one of these sessions, we were starting to discuss new product ideas, when one participant, a legal professional, before describing a thought, suddenly stopped and shook his head,…</p><blockquote><strong><em>“I’m overlawyering here, let me start again.”</em></strong></blockquote><p>We loved the sentiment behind this statement, because it emphasized so well what so many of us are too often guilty of — <em>over analyzing and getting caught in the weeds! </em>Something all of us have been guilty of at one point or another.</p><p><em>We quote this statement almost weekly.</em></p><p>As we all know, uncertainty is uncomfortable. However, uncertainty can be very useful in challenging us to think exponentially and to come up with new ideas. Going down rabbit holes of perfectionism and edge cases doesn’t help any of us in this journey.</p><p>The expectation to get to the 100% without a learning journey stifles innovation and “trying out new things”. Edge cases and small print should be for (released) products. Proof of Concepts need breathing space for experimentation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*hX52r9xWjMHF6yEGjSPZ5g.png" /></figure><h3>Safety Bias</h3><p>As knowledge workers, we wish to be 100% correct 100% of the time. We translate this desire (… and education, and training, and professional experience) into the way we think about future products and services — and it is hurting our ability to experiment and move forward quickly.</p><p><em>This behavioural pattern is called safety bias. We overweigh potential losses over potential gains.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*awjGuCKqQ-xzx2HFDR2y9g.png" /></figure><h3>Embrace the unknown</h3><p>Not knowing the right solution to a problem is painful. If we are honest, too often, we might not even know what is the right problem to solve in the first place.</p><p>In the world of software development, still, we believe in the magic of “just having the right idea. We cling on to the “first problem” and the “first idea” we identify, just to escape the uncomfortable space of not knowing. If our goal is to <em>be</em> in the state of <em>knowing</em> — we might rush to a potentially suboptimal outcome because we’re optimising for the wrong thing.</p><p><em>If we embrace the fact that we don’t know (yet) — we switch to a learning mindset.</em></p><p>Discover other, potentially better, problems to solve. Research your customers’ pain points, investigate their workflows and study and track your users’ behaviour, to detect the best opportunity. Explore a variety of ideas and allow them to fail fast if they don’t test well.</p><p><em>That’s what will lead you to the idea with the most potential.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*o_D69JyIuLj_cBZiDSfYuw.png" /></figure><h3>Aiming for 100% ?</h3><p>The moment AI drops into an innovation conversation, we often expect AI to “magically” solve all of our users and customers’ problems with a 100% accuracy.</p><p>This can tempt us to ignore the process of careful problem definition, creative problem solving and the detail of embedding AI in a valuable product.</p><p><em>AI won’t give us “100%” … and that’s a great opportunity!</em></p><p>Current AI methods are fantastic at solving specific, well-defined challenges. Being aware of the limitations of narrow AI [1] is in fact, a great place to explore collaboration and ways to enrich AI reasoning with human domain expertise.</p><p>If we start with the understanding that technology alone will not give us the perfect solution — we can design products and services that maximise human expertise.</p><p><em>A carefully designed human / machine choreography will give you the “100%”!</em></p><p>Such socio-technical systems enable human users to <em>collaborate</em> with the machine on steps such as the “interpretation of information” and the “final decision making”. It is a good idea, to design systems that bring in editorial domain expertise through easy-to-use touchpoints for customers and editorial as well as to continuously evolve the system based on user feedback and usage analytics.</p><p><strong>Read more:</strong></p><ul><li>[1] narrow or weak AI: <a href="https://en.wikipedia.org/wiki/Weak_AI#cite_note-1)">https://en.wikipedia.org/wiki/Weak_AI</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*kidLQLM_S1aUtP2NDiDR8Q.png" /></figure><h3>Experimentation Mindset</h3><p>If we manage not to “overlawyer” — we can find ways to publish experimental Proof of Concepts and beta versions earlier to try out new features in live case studies.</p><p>It’s no news that big tech shifted to a more design-led mindset, e.g. IBM [2] and SAP [3]. The legal startup ecosystem and law firms have done great deal to bring Design Thinking to their core business, e.g. Legal Geek [4] and DLA Piper [5]. Service Design [6] and Design Thinking [7] provide a wealth of methods for culture change and customer-centric product development and business design.</p><p><strong>Read more</strong></p><ul><li>[2] IBM and Design Thinking <a href="https://www.wired.com/2016/01/ibms-got-a-plan-to-bring-design-thinking-to-big-business/">https://www.wired.com/2016/01/ibms-got-a-plan-to-bring-design-thinking-to-big-business/</a></li><li>[3] SAP and Design Thinking, <a href="https://experience.sap.com/skillup/introduction-to-design-thinking/">https://experience.sap.com/skillup/introduction-to-design-thinking/</a></li><li>[4] Legal Design @Legal Geek <a href="https://www.legalgeek.co/learn/legal-design-wtf/">https://www.legalgeek.co/learn/legal-design-wtf/</a></li><li>[5] DLA Design Thinking Training <a href="https://www.thelawyer.com/dla-trains-500-employees-for-client-service-post-lockdown/">https://www.thelawyer.com/dla-trains-500-employees-for-client-service-post-lockdown/</a></li><li>[6] What is Service Design. <a href="https://www.nngroup.com/articles/service-design-101/">https://www.nngroup.com/articles/service-design-101/</a></li><li>[7] Why Design Thinking Works <a href="https://hbr.org/2018/09/why-design-thinking-works">https://hbr.org/2018/09/why-design-thinking-works</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*p1Sy7fvpCMa4SkIM08bMDQ.png" /></figure><h3>Fail (aka Learn) Fast!</h3><p><em>There is a big difference between running fast, and running in the right direction!</em></p><p>Trust the process! There are proven methods socialised in the Design Thinking and Innovation community to guide experimentation. Schools of thought such as Agile, Lean UX or Design Thinking provide plenty of ideas how to make this happen! [8][9]</p><p>Shift your focus from “deliverables”, for example, the number of features, to ”enabling outcomes”, such as “ease-of-use” or “increased sign ups”. Build incrementally on a “firm foundation” [10] and through a detailed understanding of the problem. Collaborate closely and test early with customers through co-design-partnerships.</p><p>Creative problem solving techniques [11] can help to think outside of the box. Instead of efficiency gain for existing processes, <em>critically re-think the process, explore new business models</em><strong> </strong>and different ways to achieve the desired outcome.</p><p><em>Don’t overlawyer — enable your team to experiment and fail fast.</em></p><p><strong>Read more:</strong></p><ul><li>[8] What does Agile mean? <a href="https://apmg-international.com/article/what-does-agile-mean">https://apmg-international.com/article/what-does-agile-mean</a></li><li>[9] Lean and Agile UX <a href="https://medium.com/@beartalk/lets-do-the-ux-in-a-lean-and-agile-way-lean-ux-by-jeff-gothelf-and-josh-seiden-eada9f12f78e">https://medium.com/@beartalk/lets-do-the-ux-in-a-lean-and-agile-way-lean-ux-by-jeff-gothelf-and-josh-seiden-eada9f12f78e</a></li><li>[10] Build incrementally from firm foundations: <a href="https://agilepm.wiki/dsdm-agile-project-framework/principles/build-incrementally-from-firm-foundations/">https://agilepm.wiki/dsdm-agile-project-framework/principles/build-incrementally-from-firm-foundations/</a></li><li>[11] Ideas for creative problem solving <a href="https://gamestorming.com/">https://gamestorming.com</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/50/1*GXyTF8wSXMDNREH_IbyhjA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1b12c77258b8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to: Remote design workshops]]></title>
            <link>https://uxdesign.cc/how-to-remote-design-workshops-19e47f37cb12?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/19e47f37cb12</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[remote-work]]></category>
            <category><![CDATA[design-thinking]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Thu, 18 Jun 2020 15:57:13 GMT</pubDate>
            <atom:updated>2020-06-23T08:45:05.450Z</atom:updated>
            <content:encoded><![CDATA[<h3>How to: Remote UX Workshops</h3><h4><strong>You want to run a remote Design Workshop with loads of participants online? No problem, with the right agenda and tools!</strong></h4><figure><img alt="a magnifying glass, search, a web thing, tools to help us work remotely" src="https://cdn-images-1.medium.com/max/100/1*QhYKPR8AENfJC_AGhd9bDQ.png" /></figure><p>The current situation forces us to work from home, and adapt the way we work. Yes, in-person workshops are easier and more productive. Nothing beats a collaborative whiteboarding session or paper prototyping to express and discuss some ideas. But hey, let’s see how we can move the best bits online!</p><p>This article contains some tips and tricks that are specific to Miro, but most of the points are applicable to any remote workshop and online workshop tools like Mural, Whimsical, Lucidchart etc.</p><p>This article is co-writen with <a href="https://medium.com/u/3c4b16bb3c29">Brian Romer</a> &amp; Mary Mikhail.</p><p><strong>Content:</strong></p><ul><li>Why a workshop?</li><li>Do you have the right people in the “room”?</li><li>How to: Setup</li><li>How to: Introduction</li><li>How to: Divide and Conquer</li><li>How to: Closing</li></ul><figure><img alt="a paint workshop" src="https://cdn-images-1.medium.com/max/1024/1*VUhsgeTeHKbqG1VJrUapPw.jpeg" /><figcaption>Source: ‪unsplash.com‬⁩, Thanks to <a href="http://twitter.com/Khara0ke">@Khara0ke</a> for making this photo available freely on <a href="http://twitter.com/unsplash">@unsplash</a> 🎁</figcaption></figure><h3>Why a Workshop?</h3><h4>Purpose</h4><p>What is the purpose of your workshop? For the following we’ll assume you know more or less how Design Thinking works. We assume you have a goal for the workshop, preparation and workshop outcome are nicely embedded into a wider design process.</p><h4>Outcomes</h4><p>What kind of deliverables or outcomes you’re aiming for? Check in with your stakeholders and align the exercises and tools with the desired outcome. Often the most important — yet intangible — outcome is alignment between workshop participants.</p><figure><img alt="illustration of people in conversation" src="https://cdn-images-1.medium.com/max/100/1*uGoo9pR5Px47wEnyH2RQwQ.png" /></figure><h3>Do you have the right people in the “room”?</h3><h4>Participants</h4><p>Be clear about who should be included (and why). You want participants that can contribute — rather than observers. Invite end-users, practitioners or people with an overview on the question you want to work on — rather than their managers or supervisors, if they are curious, but don’t actually want to contribute. Keep your group small, 4–6 participants works best — or break your larger group down into smaller subgroups.</p><blockquote><strong><em>Pro tip:</em></strong><em> Look up everyone’s name, title and photo on the web or LinkedIn and set up an area on Miro that introduces everyone.</em></blockquote><h4>Time Commitment</h4><p>Explain participants how much time you’ll be asking for. Remote workshops can be more taxing than in-person workshops because it’s hard to read the room and create a sense of flow, so aim for less time than you would in person.</p><h4>Time Zones</h4><p>Find a time slot that works for everyone across time zones, with enough lead time that they can all do a bit of pre-reading and homework before the event. Describe the goals concisely in the calendar meeting.</p><h4>Accessibility</h4><p>Ask if anyone has vision, hearing or other accessibility issues that we should address. Make sure that all participants are accommodated for. There could be a channel for participants who have a disability or are more camera/mic shy. Some participants might ask questions or contribute through alternate ways such as having a designated helper to share the screen for a visual reference/troubleshooting.</p><h4>Reminder</h4><p>A few days ahead of time, send everyone an email asking them to do some basic Miro tasks like making post-it notes etc. The day before, update the calendar event to remind people to do the Miro homework if they haven’t yet. You want to have all your participants familiar with using the tool before starting the workshop, to make the most of everyone’s time.</p><figure><img alt="illustration of stationary" src="https://cdn-images-1.medium.com/max/100/1*x2m62HngbNTQCJH1e2VsZQ.png" /></figure><h3>How to: Setup?</h3><h4>Shared Canvas</h4><p>You need a shared space to work collaboratively. Miro, Lucidchart, Stormboard and Mural are among the alternatives.</p><h4>Teams</h4><p>Use Teams for the audio, video and screenshare features.</p><blockquote><strong><em>Pro tip: </em></strong><em>Teams works great but when everyone is on the Miro board, you no longer see the participants’ video squares. Miro has a screenshare option that can be used along with video call functionality. Choose to stick to what your participants would feel comfortable with.</em></blockquote><h4>Tool Overload</h4><p>Try make the right trade offs on what tools you are using. Try stay in the same platforms as much as possible, e.g. choose to do the ice breaker in the same tool that you’d use for workshop.</p><h4>Breakout Rooms</h4><p>Workshops live from small side-conversations, you need separate video chat rooms. Zoom has a feature that allows you to do that pretty well, with other tools you might get a workaround by starting multiple meetings at once and direct your participants to which ones they should join.</p><h4>Work Groups</h4><p>Making the best use of everyones’ time, it’s best to split up larger groups into smaller groups. Define your work groups prior to the workshop, so that everyone knows which group they belong to.</p><h4>Rehearse</h4><p>If leading the workshop with more than one facilitator, choreograph and rehearse who speaks and screen shares when.</p><blockquote><strong><em>Pro tip:</em></strong><em> we used the notes section in the Miro board to plan out who is doing which part of the board and when they are. While working on the board, it’s where we added to-do lists for collaborative editing as well.</em></blockquote><h4>Timer</h4><p>Make sure you have a physical or digital countdown timer or alarm clock at hand. Keeping tasks time-boxed, even if a bit arbitrarily, helps to keep momentum and to preserve your agenda from running over.</p><blockquote><strong><em>Pro tip:</em></strong><em> Miro provides a great in-app countdown timer.</em></blockquote><h4>What to do?</h4><p>Draft a minute by minute agenda using sessionlab.com or similar. Identify parts of your agenda that could be discarded if time is short. Identify parts that are crucial to keep.</p><figure><img alt="example time table, 30 minutes introduction, 30 minutes setting the scene, break, 45 minutes group work, 15 minutes pitch" src="https://cdn-images-1.medium.com/max/1024/1*tMnAK3R-c4SD-dNFcF8_nQ.png" /></figure><blockquote><strong><em>Pro tip:</em></strong><em> When planning Miro workshops with participants whose comfort level is unknown to you or if their familiarity with the tool is low, budget double the amount for that person to get an average sense for the activity. Be very generous with time allocations.</em></blockquote><blockquote><em>Consider the number of participants (and their coordination). Less experienced participants might take up to 2x the time to complete a task.</em></blockquote><h4>Layout</h4><p>In the same way you would lay out exercises in a workshop room, prepare a canvas with areas for specific exercises. It might make sense to have areas for the entire group, for specific work groups and individual thinking time.</p><blockquote><strong><em>Pro tip:</em></strong><em> Structure the boards in a way that accommodates your workshop type and participants’ familiarity with Miro. If your workshop is a design sprint (where the work is done in stages), you will naturally want everyone to be on the “same page”. Whereas for more individual work, it makes sense to keep work spaces separate to keep participants focused on their work — rather than having a view of others’ workspaces. Having individual spaces for participants can be daunting for navigating. Consider how you’d like to conduct analysis as this may help shape your structure. Other things that are important to consider is sizing and layout as you will want minimal scrolling for your participants and for yourself.</em></blockquote><figure><img alt="illustration of documents and search" src="https://cdn-images-1.medium.com/max/100/1*sYQGiuXaaaQkhwCrkA7w3A.png" /></figure><h3>How to: Introduction</h3><h4>Time for technical setup</h4><p>Will there ever be a meeting that doesn’t start with 5–10 minutes of technical difficulties to get everything set up? I wish there was, until then, plan some time to get set up.</p><blockquote><strong><em>Pro tip:</em></strong><em> Consider sending the Miro link to participants along with some time in your calendar for technical office hours, where you can guide and direct them in a personable setting.</em></blockquote><h4>Ice breaker</h4><p>Start with something fun. Participants can be nervous, feel under pressure, think they are important. None of which will help you run a good session. Start with an icebreaker that slightly nudges everyone out of their comfort zone … and most importantly makes them laugh!</p><blockquote><strong><em>Pro Tip:</em></strong><em> Good ice breakers are …</em></blockquote><blockquote>Draw a picture of your each other (if participants are on video call) … without looking at your paper</blockquote><blockquote>Draw a device for / or with … a chicken / a horse / a bird</blockquote><blockquote>Draw your problem / Describe the problem through some digital post its</blockquote><blockquote>Fill in the blanks, e.g. “On the weekend I had loads of fun … “, “This week’s happy thought is … “ — it may be worthy to consider doing the ice breaker in the same tool to do a “technical ice-breaking”</blockquote><blockquote><em>For more ideas check out:<br></em><a href="https://gamestorming.com/">Gamestorming</a> <a href="https://playtaboo.com/playpage">https://playtaboo.com/playpage<br>https://playtaboo.com/playpage</a><br><a href="https://miro.com/templates/ice-breaker/">https://miro.com/templates/ice-breaker/https://miro.com/templates/ice-breaker/</a></blockquote><h4>Make it personal</h4><p>Ask participants to switch on the camera — if they are comfortable with it — its just more fun to see each other laughing, struggling and coming up with thoughts and ideas!</p><blockquote><strong><em>Pro tip:</em></strong><em> It’s really hard to gauge the room and make it light when everyone is on mute. Ask your participants to keep their audio on and to engage freely to lighten it up. You can even have participants mute and play their own music while getting creative.</em></blockquote><h4>Introduction (Main Meeting Room)</h4><p>Tell everyone why they are are, what they are going to do and what it will lead to. People are used to meetings to … talk and make decisions. The point of this session is not to come to a premature decision following the loudest voice, but to explore options and think through alternatives.</p><blockquote><strong><em>Pro tip:</em></strong><em> Use a landing page as an overview on the board. This is what participants see when they come into the workshop. If lots of placeholders are on the Miro boards, you can overlay a solid rectangle on top to hide them until you are ready to reveal the frames as you go through the workshop. Another pro-tip is to use a tool familiar to participants to initiate it. If everyone populates excel cells, they get pasted as post-its in Miro and you can have the note keeper annotate the board before the main activity starts. It may even be possible to conduct the entire session via screenshare and annotations in this manner. We sometimes do that during our user research sessions!</em></blockquote><h4>Roles</h4><p>Running a workshop successful requires different roles.</p><ul><li><strong>Facilitator </strong>to walk the group through a number of exercises, give people nudges and prompts to work off</li><li><strong>Timekeeper </strong>to keep momentum and enable the group to get through the agenda</li><li><strong>Note keeper </strong>to keep observations, quotes and learnings</li><li><strong>Technical troubleshooting</strong> to help people use the technology</li></ul><p>Facilitator and time keeper can be the same person. It is hard to effectively take notes and facilitate at the same time.</p><blockquote><strong><em>Pro tip: </em></strong><em>Once you’ve laid out all your boards or frames, in Miro you can lock the background elements so it’s harder for people to accidentally move them around. Lock objects in one large selection group rather than in smaller groups, as unlocking can get confusing otherwise.</em></blockquote><blockquote><strong><em>Pro tip: </em></strong><em>Copy paste everything into a separate board in case anything gets deleted or misaligned, you’re able to quickly recover mid-session. Have a person dedicated to help out users. Sometimes it’s help with resizing or moving things over, locking frames, etc..</em></blockquote><figure><img alt="illustration of a bucket of paint" src="https://cdn-images-1.medium.com/max/100/1*MQebIuWGOh70-UyHjp14qg.png" /></figure><h3>How to: Divide and Conquer</h3><h4>Landscaping (Main Meeting Room)</h4><p>While you have the attention of the whole group set out the agenda and define the large chunks of work, individuals or individual groups can work on.</p><figure><img alt="illustration of laying out 3 steps e.g. in a user journey, in a large group" src="https://cdn-images-1.medium.com/max/974/1*A6e5T6VjZ19kAfFXvLEGlw.png" /></figure><p>If you are working on a process of a service or application, why not break it down into steps?</p><blockquote><strong><em>Pro tip: </em></strong><em>Prepare different areas on the board for work with the whole group, sub groups and individual work spaces. Make sure to lock the right frames so they don’t move around the board!</em></blockquote><h4>Breathing space</h4><p>Some people come up with great ideas in conversation, … actually most people benefit from quiet thinking time to reflect on complex situations, recollect their observations and come up with ideas.</p><p>It is always a good idea to build your session with a mix of the following:</p><ul><li><strong>Individual thinking time</strong> — writing up observations, thoughts or ideas</li><li><strong>Workgroups </strong>— share and discuss in a smaller group, giving time to every voice</li><li><strong>Plenum </strong>— present, pitch, vote in big group</li><li><strong>Facilitation</strong> — Moderation of idea collection, voting or idea grouping exercises</li></ul><h4>Exercises</h4><p>While facilitation and well-defined tasks are key to keep the energy in the room for in-person workshops. This is even more true for remote workshops.</p><ul><li><strong>Chunks of work</strong> — Describe small pieces of work, that can be done in a short amount of time</li><li><strong>Context</strong> — Put exercises into context of a wider picture or “landscape”</li><li><strong>Time Keeping </strong>— Stick to a timing per exercise — give people control — and stop people from moving around pieces when the time is up</li></ul><p>With online workshops which have a tendency to be asynchronous (what participants are seeing may be is different), documentation is important. Adding in written instructions where applicable. If using a tool, leaving in specific instructions can have the added benefit of having users complete tasks at the right pace for them. The idea behind this one is that you have no chance to overhear side conversations and to gauge the room.</p><h4>Breaks</h4><p>Make sure to schedule time for breaks. Have everyone get up from the desk — stretch their legs, take a bio break — make sure to be clear about the time when you want them back.</p><h4>Workgroups (Breakout rooms)</h4><p>Once the steps are clear, have groups break out into smaller meeting rooms, in which they can work on defined tasks in parallel.</p><p>You want to be as efficient as possible with everyone’s time. Having a large group listen to one person at a time is not good use of workshop time. Its much better to have groups of 2–3 people work through parts of a challenge — and bring them back to the larger group.</p><blockquote><strong><em>Pro tip:</em></strong><em> assign groups before hand so everyone knows in which group they are. prepare different areas on the board for each sub group. People can easily copy paste notes and ideas they come up with.</em></blockquote><p>Again, you might want to follow a breathing structure. Start with your first big item — and give people individual thinking time to reflect and write up ideas for problems or solutions.</p><figure><img alt="illustration to have different work groups focus on different steps" src="https://cdn-images-1.medium.com/max/974/1*jfk_NHkZgHe2WHE8UcrLxg.png" /></figure><p>Have them share ideas — one at a time. Work through the material together and further refine notes or group them into topics.</p><figure><img alt="illustration of up voting individual ideas per work group" src="https://cdn-images-1.medium.com/max/974/1*TWYvNLoVr1rRJNFKoqk9mQ.png" /></figure><p>Have the group vote for the most prominent pain points or the best ideas that should be moved forward and back to the main meeting room.</p><figure><img alt="illustration of bringing upvoted ideas back into the larger group" src="https://cdn-images-1.medium.com/max/974/1*bNQg-Sel0ftffzjcm2cE-g.png" /></figure><p>Give each workgroup time to further flesh out and prepare a presentation back to the larger audience.</p><h4>Combine (Meeting room)</h4><p>Getting everyone at the same time is a challenge. Maybe give early teams a coffee break — and have them start presenting once everyone is ready.</p><p>First, check in whether ideas and observations still fit into the larger landscape — or make changes if necessary.</p><p>Then have each workgroup present — on a strict timeline — that also leaves time for questions and clarifications. Have a different person serve as a note taker to fill in things while people are sharing.</p><figure><img alt="illustration of merging upvoted ideas back into the larger landscape" src="https://cdn-images-1.medium.com/max/974/1*jLAliw_9wrKTAeCBEDs5hw.png" /></figure><p>Get everyone to vote on the outcomes. Get to a consensus about priority or next steps. Add the links to video/recordings/other artifacts in in the board. While things are fresh, add any relevant detail for documentation.</p><p>When using a whiteboarding tool, going back to it for surveys/followups/comments is sometimes very easily done and and closes the feedback loop faster.</p><figure><img alt="illustration of web search" src="https://cdn-images-1.medium.com/max/100/1*i6E212b1gDX_G72YPVkgNA.png" /></figure><h3>How to: Closing the workshop</h3><p>If your exercises went as expected and you hit the time … respect! This is tough to do especially online. Plan buffer time — and a fun way to close the session</p><ul><li>You can ask everyone to draw a picture about their experience with the workshop … in 1 minute!</li><li>You can ask everyone how it went for them, about a thing they were surprised about or learned</li><li>You can share a print out of the workshop board right after the workshop — which gives you more time to actually write up learnings</li></ul><p>Did you conduct any online workshops so far? If yes, how did it go? We are also very curious to hear whether any of the above helped you run a future workshop?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/76/0*jRsx40IsMQ2JQ-ey" /><figcaption>The UX Collective donates US$1 for each article published in our platform. This story contributed to <a href="https://medium.com/uxmpretas">UX Para Minas Pretas</a> (UX For Black Women), a Brazilian organization focused on promoting equity of Black women in the tech industry through initiatives of action, empowerment, and knowledge sharing. Silence against systemic racism is not an option. Build the design community you believe in.</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=19e47f37cb12" width="1" height="1" alt=""><hr><p><a href="https://uxdesign.cc/how-to-remote-design-workshops-19e47f37cb12">How to: Remote design workshops</a> was originally published in <a href="https://uxdesign.cc/">UX Collective</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Notes on Design (UX.Live Conference)]]></title>
            <link>https://medium.com/@jshlth/notes-on-design-ux-live-conference-e84d981256d4?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/e84d981256d4</guid>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[human-ai-interaction]]></category>
            <category><![CDATA[conference]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[user-experience]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Sat, 14 Dec 2019 10:37:40 GMT</pubDate>
            <atom:updated>2019-12-14T10:40:39.514Z</atom:updated>
            <content:encoded><![CDATA[<p>The 2nd year edition of a more mature UX.Live conference invited some fantastic speakers over the course of 3 days.</p><p>This blogpost summarises my notes and thoughts concerning <strong>Design </strong>practice. It comments on talks by <a href="https://twitter.com/noel_lyons">Noel Lyons (@noel_lyons)</a>, Morgane Peng (<a href="https://medium.com/@morganepeng">https://medium.com/@morganepeng</a>), David Attwater (<a href="https://www.eiginc.com/WP/">Enterprise Integration Group</a>), <a href="https://twitter.com/rbanks">Richard Banks (@rbanks)</a>.</p><p>Make sure to check out my other article about <a href="https://medium.com/@jshlth/user-research-ux-live-conference-12247bde3774">User Research UX.Live</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*m0Yap_FgIKD3lOJ8R1gCuA.png" /></figure><h3>Design Contributions</h3><p>Often people ask about the ROI of Design — treating it like a delivery function only. Building on a talk by Noel Lyons (Barclays) and my own experience I would argue that design contributes different:</p><ul><li><strong>Empathize</strong> and answer the question what problem to solve — and <em>why</em></li><li><strong>Explore</strong> a breadth of ideas to find a more appropriate solution</li><li><strong>Illustration</strong> and vision building</li><li><strong>Collaboration</strong>, sharing of ideas and knowledge</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WVZnNxJoNnMquKGyfRstjQ.jpeg" /><figcaption>Design Support, Slide Noel Banks</figcaption></figure><p>In a fun talk Morgane Peng shared ideas how to deal with stakeholders that <em>don’t get </em>Design. Her simple, but crucial steps to a seat at the table:</p><ul><li><strong>Learn the business </strong>— and enough domain expertise to really contribute to discussions</li><li><strong>Be an ally, not a resource</strong> — and align to business strategy</li><li><strong>Be part of the solution</strong> — solve business challenges</li></ul><p><strong>Read more:</strong></p><ul><li><a href="https://uxdesign.cc/how-to-deal-with-people-who-dont-get-design-design-and-strategy-3a1ef8c1841e">How to deal with People who don’t get Design (uxdesign.cc/)</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*ruV4QAjOQ4hBOMvrDDQhfQ.png" /></figure><h3>Design for Interaction with Intelligent Systems</h3><p>Various talks touched on Design for/with AI or new technology.</p><p>David Attwater presented on design and development of Voice interaction. It brings a number of specific challenges that are also relevant for interaction with intelligent actors generally.</p><h3>Lack of a Metaphor</h3><p>Voice interfaces literally provide no metaphor — equally we tend to overestimate interaction with intelligent systems for example through keyword based search. David brilliantly summarised Theory Of Mind with a simple illustration.</p><h3>Theory of Mind</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D2IbngwOYUvAWjyCMzJaMg.jpeg" /><figcaption>I think therefore I am … Slide David Attwater</figcaption></figure><p>As an individual we constantly visualise our position in our environment (physical and social). We perceive our surroundings and their constraints on our actions.</p><p>When interacting with a cognitive machine we assume it to behave in a similar way, namely to be aware of the shared environment. We not only wonder about what it can do, what it is able to understand and remember — but also what it thinks that <em>we </em>might know. On the other hand we expect it to have some understanding and model about what the we want, prefer and have done previously</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BO4FCg9ltFkX6qijSM_7IQ.jpeg" /><figcaption>I think about, how you think about, how I think about … Slide David Attwater</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*R3YFViGcpns_CwL5Wdw50A.png" /></figure><h3>Linguistic Alignment</h3><p>During interaction through language we try to align with our conversational partner in various ways:</p><p><strong>Lexical alignment</strong>– use of the same words/terminology</p><p><strong>Grammatical alignment</strong>– use of similar grammatical structures:</p><ul><li>“What to time does your shop close?” — “five o’clock”</li><li>“<strong>At </strong>what time does your shop close?” — “<strong>At</strong>five o’clock”</li></ul><p><strong>Syntactic alignment</strong>–passive/active:</p><ul><li>“he gave the book to her” — “she gave it back to him”</li><li>“the book was given to helen” — “then it was given back”</li></ul><p><strong>Structural alignment</strong></p><ul><li>“Go one cross then three down” — “I’m at one up and four across”</li><li>“go to box A 4” — “I’m at B 6 currently”</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*ot1lAlfsZ5LunbkLki7WTA.png" /></figure><h3>Guidelines for Human-AI-Interaction</h3><p>Richard Banks spoke about designing for and with AI. Data-driven solutions often break with key usability principles, for example personalised search breaks with a search experience that is “consistent” across users and time. Different users will get different results — as will the same user at different times.</p><p>It is difficult to understand limitations and errors (any) intelligent systems make. Explanations should give an indication of constraints — and carefully reflect the granularity that can be expected from the underlying model.</p><p>Microsoft published guidelines that can be used to probe and steer design efforts. For each prompt there are obviously numerous solutions — depending on the system and context.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9H5HsCEWPQHHmfI34HQR_g.jpeg" /><figcaption>Illustration Microsoft Guidelines for Human-AI Interaction</figcaption></figure><p><strong>Read more</strong></p><ul><li><a href="https://www.microsoft.com/en-us/research/project/guidelines-for-human-ai-interaction/">https://www.microsoft.com/en-us/research/project/guidelines-for-human-ai-interaction/</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*EF0At2U8C_ELHX8OWhhnmA.png" /></figure><p>Why there are all these light bulbs and pen’s sprinkled through the post? Design is all about ideas — find the full set of (beautiful) icons here <a href="https://thenounproject.com/Graphic_Tigers/collection/web-seo-glyphs-icons/">Web &amp; Seo Glyphs Icons Collection | Noun Project</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e84d981256d4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[User Research (UX.Live Conference)]]></title>
            <link>https://medium.com/@jshlth/user-research-ux-live-conference-12247bde3774?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/12247bde3774</guid>
            <category><![CDATA[design]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Sat, 14 Dec 2019 10:35:21 GMT</pubDate>
            <atom:updated>2019-12-14T10:40:52.901Z</atom:updated>
            <content:encoded><![CDATA[<h3>Notes on User Research (UX.Live Conference)</h3><p>The 2nd year edition of a more mature <a href="https://uxliveconference.com/">UX.Live</a> conference in London invited some fantastic speakers over the course of 3 days.</p><p>This blogpost summarises my notes and thoughts concerning <strong>User Research</strong> practice. It comments on talks by Alexander Muir (<a href="https://twitter.com/Alexander108">@Alexander108</a>), Kasper Friis (<a href="https://uxplanet.org/@kafri">https://uxplanet.org/@kafri</a>), Saskia Liebenberg (<a href="https://twitter.com/SaskiaLieb">@SaskiaLieb</a>) and Etienne Fang (<a href="https://medium.com/@etiennefang">https://medium.com/@etiennefang</a>).</p><p>Make sure to check out my other post about <a href="https://medium.com/@jshlth/notes-on-design-ux-live-conference-e84d981256d4">Design (UX.Live Conference)</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*40YyIvCiY7D6tAzFcpGLTg.png" /></figure><h3>Don’t optimize — differentiate!</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BOg12t6hAYu7DfBGjLQF_w.jpeg" /><figcaption>It’s not an optimization problem — Slide, Alexander Muir</figcaption></figure><p>Alexander Muir gave an inspiring workshop on planning and executing user research. He advised for innovation projects and improvement of services and systems to not (only) look at optimization. Yes, efficiency and effectiveness (key considerations in usability and UX) are crucial but a blind push for optimization might — in the worst case — only re-enforce existing and broken processes. It’s key to rethink processes — starting from the user — and aim at differentiation in our service offering.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*oi1b5IWa8wedx4j0dwr4dA.png" /></figure><h3>Study Your Problem</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NwpmJPzURgeZrNRnr1-crw.jpeg" /><figcaption>Disruptive Design Method — Slide, Alexander Muir</figcaption></figure><p>A design inventory of existing (and competing) systems and services helps to build up a vocabulary of objects that can be involved/reassembled and build upon during the design phase.</p><p>A neat framework to structure the approach is provided by <strong>Disruptive Design Method</strong>. It outlines three phases:</p><ul><li><strong>Problem loving</strong> — study and mine your problem and deeply empathize with the humans involved</li><li><strong>Systems Mapping</strong> — laying out the landscape, actors and datascape of a problem and its context, desk research on reports and terminology involved in a subject matter</li><li><strong>Ideation &amp; intervention</strong> — generation of various new ideas, testing and implementation</li></ul><p><strong>Read more:</strong></p><ol><li><a href="https://www.disruptdesign.co/the-disruptive-design-method">https://www.disruptdesign.co/the-disruptive-design-methodThe Disruptive Design Method — DISRUPT DESIGN</a> <a href="https://www.disruptdesign.co/the-disruptive-design-method">https://www.disruptdesign.co/the-disruptive-design-method</a></li><li><a href="https://vimeo.com/219482990">https://vimeo.com/219482990The Disruptive Design Method by UN Champion of the Earth Leyla Acaroglu on Vimeo</a> <a href="https://vimeo.com/219482990">https://vimeo.com/219482990</a></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/728/0*LqLyMctg71OX3u1q.gif" /><figcaption>Being lost … <a href="https://uxdesign.cc/ux-metric-lostness-524998ccb215">UX Metric Lostness</a> (<a href="https://uxdesign.cc/ux-metric-lostness-524998ccb215">uxdesign.cc</a>)</figcaption></figure><h3>Lostness</h3><p>Net Promoter Score (NPS) and Customer Satisfaction (CSAT) are not great to actually inform how to improve a service. Neither capture the full picture when used in isolation. Alexander Muir spoke a refresher on a metric that has been around for a while — but is often ignored: <strong>Lostness, </strong>is a useful measure to evaluate navigation support and orientation in a user experience adding to the mix of classical usability metrics (Efficiency, effectiveness, satisfaction etc.)</p><p><strong>Read more:</strong></p><ul><li><a href="https://uxdesign.cc/ux-metric-lostness-524998ccb215">UX Metric Lostness</a> (<a href="https://uxdesign.cc/ux-metric-lostness-524998ccb215">uxdesign.cc</a>)</li><li><a href="https://www.nngroup.com/articles/nps-ux/">Net Promoter Score: What a Customer-Relations Metric Can Tell You About Your User Experience</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*bwtCS2Dc8xHXXk9NMhTSwg.png" /></figure><h3>“Strong Opinions Lightly Held”</h3><p>All innovation starts with some idea — and inherent bias. A useful notion how to deal with that, is the idea to articulate our opinions — hopes and fears — but encourage the team to overthrow these assumptions as soon as new research indicates otherwise. While user research (both quant and qual) aims for solid, reliable insight — it’s never going to be perfect. There is always the need for a user researcher to take a risk, decide on key observations and their interpretation.</p><h3>“Trim Tabs”</h3><p>The notion of small changes that hugely steer experience and outcome is particularly useful to articulate why a fair amount of research might only lead to a small (but impactful) change.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/220/1*a_CoqU6O_kXUsWMGu1yWGQ.png" /><figcaption>Trim Tabs (wikipedia)</figcaption></figure><h3>One thing a week + Layering Studies</h3><p>Good advice for small design teams or smallish projects — was to continuously do one piece of user research a week, in order keep everyone in an experimenting mindset where testing and evaluation is a default — rather than an exception. Keeping — even limited — user research actionable is key — especially when working in agile teams. It is better to deliver 80% of something on time, rather than 100% of nothing. When planned carefully — many small pieces of interviews and user testing can be combined to build a more comprehensive picture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BwbcH5kEO2r1omzI2hRn-g.jpeg" /><figcaption>Prioritise User Research, Slide Alexander Muir</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2OFjMRv4E_fZX38NPbMELw.jpeg" /><figcaption>Layer studies, Slide Alexander Muir</figcaption></figure><h3>Quantifying User Experience</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/950/1*7GIYYNXcsF9Oex_Btc8rhA.jpeg" /><figcaption>Illustration, Wine Taste Experience</figcaption></figure><p>User Experience is difficult to capture and quantify. Kasper Friis presented a research method using multidimensional scaling to make sense of similar — and different — experiences. This is commonly used in market research, but rarely made its way into UX, due to the required large sample size. Similar to Microsoft Reaction Cards, the method tries to describe experience through a bag of attributes — each ranked on a Likert scale. Projecting these different ratings — from multi dimensional to a 2-dimensional plot — can help elicit key attributes that made some experiences stand out.</p><p><strong>Read More</strong></p><ul><li><a href="https://www.quirks.com/articles/don-t-forget-the-graphic-multidimensional-scaling">https://www.quirks.com/articles/don-t-forget-the-graphic-multidimensional-scalingFind Market Research Companies, Facilities, Jobs, Articles, More | Quirks.com</a> <a href="https://www.quirks.com/articles/don-t-forget-the-graphic-multidimensional-scaling">https://www.quirks.com/articles/don-t-forget-the-graphic-multidimensional-scaling</a></li><li>The method could be applied to evaluate and compare even the smallest aspect of an experience, e.g. which car door sounds the most robust or reliable <a href="https://www.youtube.com/watch?v=y9Ww5W26MTE">https://www.youtube.com/watch?v=y9Ww5W26MTE</a> — or experience <a href="https://www.youtube.com/watch?v=R2IUwElbCnc">https://www.youtube.com/watch?v=R2IUwElbCnc</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*ER7q6rQBWcnvVQ0IxiYhAw.png" /></figure><h3>ResearchOps, Knowledge Management + Kaleidoscope</h3><p>Saskia Liebenberg presented on #ResearchOps which — similar to more known TechOps — is all about organisation, logistics and management of user research. Even for small design and research teams it might be useful to refer to — and contribute to — best practices laid out on <a href="http://researchops.community/">http://researchops.community</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pG5zje6Pw7GKu7QZ39xxmQ.jpeg" /><figcaption>Insight Statement, Slide Etienne Fang</figcaption></figure><p>Etienne Fang gave an inspiring talk about the insights platform Kaleidoscope at Uber that aims to collect learnings across the globe. It’s essentially a knowledge management platform to capture and organise any discovery about how users interact with a service. Each note is structured into:</p><ul><li><strong>Who</strong> — which user was involved</li><li><strong>Where</strong> — geography and time</li><li><strong>What</strong> — observed situation</li><li><strong>Why</strong> — assumed explanation of the observation</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-_ssvMhXulSQWe6wOevGgA.jpeg" /><figcaption>Insight Statement, Slide Etienne Fang</figcaption></figure><p><strong>Read more:</strong></p><ul><li><a href="http://researchops.community/">http://researchops.community</a></li><li><a href="https://airtable.com/shrWD5GpnFCPNDlKR/tbloTst2aPNRbAwXP/viwhQGhPaUl14TloW?blocks=hide">https://airtable.com/shrWD5GpnFCPNDlKR/tbloTst2aPNRbAwXP/viwhQGhPaUl14TloW?blocks=hide</a></li><li><a href="https://medium.com/uber-design/the-power-of-insights-a-behind-the-scenes-look-at-the-new-insights-platform-at-uber-26f85becc2e6">https://medium.com/uber-design/the-power-of-insights-a-behind-the-scenes-look-at-the-new-insights-platform-at-uber-26f85becc2e6</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*H65oedXOdQie8JVZrCYq2w.png" /></figure><p>Why there are all these people sprinkled through the post? User Research is all about people — find the full set of (beautiful) icons here <a href="https://thenounproject.com/marialuisa.iborra/collection/people-2d/">People 2d Collection | Noun Project</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=12247bde3774" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Collective [Human & Machine] Intelligence]]></title>
            <link>https://medium.com/@jshlth/collective-human-machine-intelligence-35874bc27450?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/35874bc27450</guid>
            <category><![CDATA[collective-intelligence]]></category>
            <category><![CDATA[service-design]]></category>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[system-thinking]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Fri, 25 Oct 2019 09:59:33 GMT</pubDate>
            <atom:updated>2019-10-25T10:01:43.462Z</atom:updated>
            <content:encoded><![CDATA[<p>Conference Notes, 21st century common sense: Using collective intelligence to tackle complex social challenges, Oct 16 2019</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*qSQB_PqFJvKJwlAVe_j-Qg.png" /></figure><h3>What’s more intelligent, people or machines?</h3><p>Don’t get me wrong, machine learning and artificial intelligence are fascinating. A combination of statistical, computational and mathematical methods — able to identify patterns, simulate understanding and even transfer learning from one domain to another.</p><p>… but so are people! Humans are just really good at having a life, lived experience, domain expertise, being part of the world and knowing about context. I would argue that what makes any system intelligent is often somehow a human element. Engineers that fine tune and maintain a system, a power user and domain experts that feed expertise into the system, or contributors who act as eyes and ears or react to a system.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*7qNxj8lqKDaOEWVHuiTI9g.png" /></figure><h3>What if … we wouldn’t have to choose?</h3><blockquote>“A simple recurrent question: How can we mobilise more intelligence — data, insights, ideas — to solve problems?”<em>- Geoff Mulgan (Chief Executive, NESTA)</em></blockquote><p><em>Collective Intelligence</em> (CI) describe the idea to combine human and machine intelligence and consider both when designing systems that take into consideration context, prior experience and the ability to abstract, unlearn, relearn and react. By keeping the human-in-the-loop with computational systems, we can use artificial intelligence to facilitate “wisdom of the crowds” and design systems that scale beyond what is possible by either humans or machines only.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*AtZSMQVC7SFezA6GcCpOXQ.png" /><figcaption>Collective Intelligence Playbook, NESTA</figcaption></figure><p>According to <a href="https://www.nesta.org.uk/team/aleks-berditchevskaia/">Aleks Berditchevskaia</a>and <a href="https://www.nesta.org.uk/team/peter-baeck/">Peter Baeck</a> (both at NESTA) CI-based approaches provide new opportunities in the following ways:</p><ul><li><strong>Understand Problems </strong>Generate contextualised insights, facts and information on the dynamics of a situation.</li><li><strong>Seek Solutions</strong>Find novel approaches or tested solutions from elsewhere. Or incentivise innovators to create new ways of tackling the problem.</li><li><strong>Decide and Act</strong>Make decisions with, or informed by, collaborative input from a wide range of people and/or relevant experts.</li><li><strong>Learn and Adapt</strong>Monitor the implementation of initiatives by involving citizens in generating data, and share knowledge to improve the ability of others.</li></ul><p><strong>Read more:</strong></p><ul><li><a href="https://www.nesta.org.uk/blog/aici-future-minds-and-machines/">https://www.nesta.org.uk/blog/aici-future-minds-and-machines/</a></li><li><a href="https://www.nesta.org.uk/toolkit/collective-intelligence-design-playbook/">https://www.nesta.org.uk/toolkit/collective-intelligence-design-playbook/</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*dCgIhP0z6Aeli6aclYZ5yg.png" /></figure><h3>Collective Crowdsourcing</h3><p>Last week NESTA’s Centre for Collective Intelligence brought together researchers from various disciplines to share ideas and findings in London, UK. Panel discussions and talks shared a number of projects that apply CI to solve system challenges.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*7UUSIdrIBXNJAcKTjIgjdQ.jpeg" /><figcaption>Illustration, Breadline</figcaption></figure><p><strong>Breadline</strong> | NGO volunteers use an interactive collaborative map to find left over goods from bakeries to plan an optimal route and collect food for redistribution. The found Daisy Tam is researching food security and the use of such tools to alleviate its challenges.</p><p><strong>Read more:</strong></p><ul><li><a href="https://breadline.hkfoodworks.com/">https://breadline.hkfoodworks.com</a></li><li><a href="https://www.youtube.com/watch?v=2p5hjUAh3CI">https://www.youtube.com/watch?v=2p5hjUAh3CI</a></li><li><a href="https://hk.asiatatler.com/society/eco-heroes-daisy-tam-breadline-hong-kong-foodworks">https://hk.asiatatler.com/society/eco-heroes-daisy-tam-breadline-hong-kong-foodworks</a></li></ul><p><strong>OSR4Rights | </strong>In this research project led by Yvonne McDermott Rees (Swansee University) human rights violations are documented through collective intelligence. Huge volumes of recordings on mobile phones allow to collect evidence, yet identification of violations is challenging.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*9v3_Q3cn1XbPiGh9DgyTqQ.png" /><figcaption>Illustratio nOSR4Rights, Source: nesta.org.uk</figcaption></figure><p><strong>Read more: </strong><a href="https://osr4rights.org/">https://osr4rights.org</a></p><p>Many more case studies can be found here: <a href="https://www.nesta.org.uk/feature/ai-and-collective-intelligence-case-studies/">https://www.nesta.org.uk/feature/ai-and-collective-intelligence-case-studies/</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*LWWIM39YIXhlJHPSFT17Cg.png" /></figure><p>A panel on <strong>designing AI that extends human capabilities</strong>brought up some interesting thoughts.</p><p><strong>Authorship</strong> | Deep learning “makes all past authors part of the collective intelligence [or authorship]” says Marco Marchesi referring to art generate based on historical art data sets.</p><p><strong>Triage | </strong>Julien Cornebise talked through case studies in health or emergency situations where automated assistants can help triage urgent cases from less urgent cases — but leave the actual assessment and decision making to the human expert</p><p><strong>Read more: </strong><a href="http://www.cornebise.com/julien/">http://www.cornebise.com/julien/</a></p><p><strong>Boredom and creativity </strong>| Karina Vold argued that by striving for ever more efficiency and by removing “boring” and basic repetitive tasks, we potentially also eliminate time to think, reflect and serendipitously discover short cuts.</p><p><strong>Read more: </strong><a href="https://www.kkvd.com/">https://www.kkvd.comhttps://www.kkvd.com</a></p><p><strong>Bias </strong>| A brilliant example going “to far” was presented in the “Is There a perfect mum” advertisement. The deliberate campain showed a portrait of an “ideal” mum based on a biased data set, containing blonde, white female models only.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fblGtBM1BaKyw76ctSAtVw.png" /><figcaption>“Is there a perfect mum?”, source: HappyFinish</figcaption></figure><p><a href="https://www.happyfinish.com/projects/baby-dove-perfect-mum-ai/">https://www.happyfinish.com/projects/baby-dove-perfect-mum-ai/</a></p><p><strong>Alternative Data Sources </strong>| In the attempt to tackle emergency situations such as Forest and peatland fires in data poor areas Maesy Angeline (Pulse Lab Jakarta) described the approach to use mobile phone connection data or social media (e.g. instagram) in combination ethnographers on the ground to better understand the impact of the disaster.</p><p><strong>Read more:</strong></p><ul><li><a href="https://youtu.be/pR4QVX_eisI">https://youtu.be/pR4QVX_eisI</a>, Haze Gazer, Pulse Lab Jakarta</li><li><a href="https://www.pulselabjakarta.org/">https://www.pulselabjakarta.org</a></li><li><a href="https://medium.com/pulse-lab-jakarta/haze-gazer-a-crisis-analysis-tool-23d7333c657a">https://medium.com/pulse-lab-jakarta/haze-gazer-a-crisis-analysis-tool-23d7333c657a</a>I</li></ul><p><strong>Creative Play</strong> | Artists showcased their studios’ work on engaging with space and artificial objects (Florian Ortkrass, <a href="https://www.random-international.com/">RANDOM INTERNATIONAL</a>), engaging with public space and cities (Usman Haque, <a href="https://umbrellium.co.uk/">https://umbrellium.co.uk</a>) and 3D representations (Kadine James, <a href="http://www.hobs3d.com/">www.hobs3d.com</a>). A fun case study showed how kids could interact with a Building Management System and learn about the status and sensors of their college building.</p><p><strong>Read more:</strong></p><ul><li><a href="https://youtu.be/Ik2bogIsuUQ">https://youtu.be/Ik2bogIsuUQ</a>, Augmented BMS, Umbrellium</li><li><a href="https://umbrellium.co.uk/products/augmented-bms/">https://umbrellium.co.uk/products/augmented-bms/</a></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*dCgIhP0z6Aeli6aclYZ5yg.png" /></figure><h3>What’s in for us?</h3><p>Collective Intelligence, Systems Thinking, Service Design, all these movements have one idea in common. That is, to design computational touch points as well as to consider human expertise and define human roles that input in or react as core part of the system.</p><p>The tools we provide are never used in isolation. They are always part of an ecosystem, used in combination with various other applications and services and based some output or used in order to input to another system. At the same time, many of the services are provided based on human domain experts input. Being more aware of this emerging ‘Collective Intelligence’ certainly could help us make even better product decisions!</p><h3>Getting Started …</h3><p>Whenever thinking about an intervention (PoC, new service, process change etc.) we should ask ourselves the following questions:</p><ul><li>What are all factors that contribute to a system’s intelligence?</li><li>Which (machine) team (algorithm working with data experience) are we putting on which tasks?</li><li>What data set, aka experience, are we hiring?</li><li>What are biases within that experience?</li><li>Which tools and algorithms are we hiring?</li><li>Which (human) team are we putting on which tasks?</li><li>Who contributes content?</li><li>Who contributes evaluation/rating/training?</li><li>Who reacts to other actors output?</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/100/1*dCgIhP0z6Aeli6aclYZ5yg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=35874bc27450" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Who Owns the Future?]]></title>
            <link>https://medium.com/@jshlth/who-owns-the-future-ad5bc11eb944?source=rss-9cd1af07b838------2</link>
            <guid isPermaLink="false">https://medium.com/p/ad5bc11eb944</guid>
            <category><![CDATA[user-experience]]></category>
            <category><![CDATA[human-in-the-loop]]></category>
            <category><![CDATA[conference]]></category>
            <category><![CDATA[future]]></category>
            <category><![CDATA[futurefest]]></category>
            <dc:creator><![CDATA[Johannes Schleith]]></dc:creator>
            <pubDate>Tue, 14 Aug 2018 18:24:53 GMT</pubDate>
            <atom:updated>2018-08-14T19:24:51.619Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*otX5TrHspIei1tCEK8Qv0w.jpeg" /><figcaption>FutureFest 2018</figcaption></figure><p><em>Who owns infrastructure, data and access to technology and therefore opportunity? How to create an equal future, nurture diversity and enable ownership — rather than consumerism?</em></p><p>Celebrating the future at Nesta’s <a href="http://futurefest.org/">FutureFest</a> in steaming hot Tobacco Dock in East London for two days in July challenged our status quo and reflected on some of these questions. Below just some observations and questions for the conference.</p><h3>Unequal futures? What about the present?</h3><p>Debates were kicked of by brilliant panel discussion by <a href="https://twitter.com/profkepickett">Kate Pickett (@ProfKEPickett)</a>, <a href="https://twitter.com/ProfSteveFuller">Steve Fuller (@ProfSteveFuller)</a> and <a href="https://twitter.com/Kat_Arney">Kat Arney (@Kat_Arney)</a> on need for equality in research and technology. It was stated that (academic and commercial) research and development of new technology seems to focus on supporting problems of elites primarily rather the wider public. While future genetic enhancements are hyped for example, too little resource is put on today’s challenges such as chronic stress or overweight.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*zsbpGKMkl1SS9Txd_n-HZg.jpeg" /><figcaption><strong>Nemesis Machine </strong>— a Smart city sculpture — <a href="https://www.futurefest.org/experiences/nemesis-machine">https://www.futurefest.org/experiences/nemesis-machine</a></figcaption></figure><p>The discussion reminded me a lot on the current discussion around artificial intelligence. We love new, smart solutions and technology! .. and while we desire a magical new thing that is smart, beautiful and solves all our problems, we too often de-prioritize less tangible, underlying behavioural challenges. It simply doesn’t sound exciting to solve basic (human) challenges in workflows or process, that involve people, rather than artefacts.</p><h3>We need diversity now!</h3><p>Discussing diversity and access to technology in the context of a FutureFest is ironic. We need diversity now! <a href="https://twitter.com/y_ychauhan">Yogesh Chauhan (@y_ychauhan) </a>and <a href="https://twitter.com/aimafidon">Dr Anne-Marie Imafidon MBE (@aimafidon) </a>discussed the need for female role models, a more diverse narrative and actual pathways in society to promote diverse work force, inventors and entrepreneurs.</p><p>Absolutely! … and I would like to freely expand this with the need to attract and educate more creative minds that think differently and bring in alternative perspectives. Research and technology has so much unrealised opportunity for those that challenge and re-think the status quo and add a different perspective. We could call this diversity, as well as multi-disciplinary and cross-functional.</p><h3>Culture | <strong>drives</strong> | Business Models | informs | Research</h3><p><a href="https://twitter.com/evgenymorozov">Evgeny Morozov (@evgenymorozov)</a> compared different models and motivations for research and development of technology — a liberal, business-model-driven approach in the silicon valley and much more state controlled approach in china, while Europe lacks a common strategy to incentivise research.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*2ybhy3OzKNYmVIXmMKCpQQ.jpeg" /><figcaption><strong>The Garden</strong> — an underground garden <a href="https://www.futurefest.org/experiences/garden">https://www.futurefest.org/experiences/garden</a></figcaption></figure><p>Speaking as a European, just as a thought experiment, a European Uber, Amazon or Alibaba might look very differently to the west coast version. Different societies have a different understanding of privacy, data protection, diversity, freedom of speech, political discourse and social welfare — all of which can impact design and product definition. Europe by nature has always been diverse in languages, cultures and political opinions — which can only benefit creativity and innovation in my opinion.</p><h3>Collaboration vs competition with the machine</h3><p>Lets keep the human-in-the-Loop and combine Collective Intelligence with Artificial Intelligence to create truly collaborative systems that help all contributors, stakeholders and ideally wider public to understand our complex problems. In his presentation on Nesta’s new effort to set up <a href="https://www.nesta.org.uk/project/centre-collective-intelligence-design/">Centre for Collective Intelligence</a>, <a href="https://twitter.com/geoffmulgan">Geoff Mulgan (@geoffmulgan)</a> out forward this focus on collaboration vs competition with the machine.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TueroLSYFjsggs3gGl0a0w.jpeg" /><figcaption>Source: <a href="https://twitter.com/KateNicolson/status/1015576988806012928">Kate Nicolson: “#FutureFest18 learning loops in collaborative intelligence, AI has loop 1 support but not 2 a…</a></figcaption></figure><p>Geoff described 3 levels of learning loops for intelligent reasoning:</p><ul><li><strong>1st Loop:</strong> Adapting thought and action within an existing framework. Huge investment in AI development are put into this 1st loop, which is all about automation and optimisation of reasoning on a given set of data.</li><li><strong>2nd Loop:</strong> Creating new categories and models to think with. This is basically what happens when we learn and adopt new terms and methods. Its much harder for for an AI to actually rethink.</li><li><strong>3rd Loop:</strong> Re-thinking how to think. Innovation, disruption and anything that changes our mental model and therefore behaviour, and ultimately really hard to put into an AI — if possible at all?</li></ul><p>In my opinion this is a good model to think about AI, data science innovation and why we need a human-in-the-loop more often than we think! This is also very relevant for TRLabs work in legal, financial and tax &amp; accounting innovation.</p><h3>Own the Future!</h3><p>Despite — or maybe because of — cynicism, criticism and debate the conference left me with a postive urge to act. While the panel discussions and talks provided loads of food for thought — very few ideas were put forward. I feel it would be great to involve more practitioners, tech companies, technologists and designers in the discussion!</p><p>One great example of a more playful, creative yet practical reflection on technology and future was presented by <a href="https://twitter.com/kcimc">Kyle McDonald (@kcimc)</a>. He showed some inspiring work, e.g. <a href="http://www.exhaustingacrowd.com/birmingham">EXHAUSTING A CROWD</a> — more to be found here <a href="http://www.kylemcdonald.net/">Kyle McDonald.net</a>.</p><p>Back in the day Bauhaus shaped the idea to design products that are easy to use, beautiful and most importantly affordable. In light of the questions of access, diversity and ownership, I am wondering how todays tech scene compares to such an approach. Wouldn&#39;t it be wonderful to create tools that not only attract paying customers, but offer similar capabilities to support the wider public and open data to foster innovation?</p><p>Another idea described in <a href="https://medium.com/@wordsandsuch/futurefest-2018-told-us-to-fear-the-future-rather-than-be-hopeful-for-it-e1b85c6a8bf6">futurefest-2018 told us to fear the future rather than be hopeful for it</a> is to describe possible future scenarios in conceptual designs — to help think about and shape concrete goals rather than abstract discussions. An image shows more than a thousand words. The same, I dare to say, is true for a good service blueprint, a sound concept or even a future narrative. If anything we need more of such Speculative Design (<a href="http://www.dunneandraby.co.uk/content/books/690/0">Dunne &amp; Raby</a> or <a href="https://www.nesta.org.uk/blog/speculative-design-a-design-niche-or-a-new-tool-for-government-innovation/">Nesta.org.uk — Speculative Design</a>) and positive, human narratives!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ad5bc11eb944" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>