<?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 CaiCai on Medium]]></title>
        <description><![CDATA[Stories by CaiCai on Medium]]></description>
        <link>https://medium.com/@hi_caicai?source=rss-c2a2cc14885d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*4MeXgDSvcGRzpWzPBCPRlQ.png</url>
            <title>Stories by CaiCai on Medium</title>
            <link>https://medium.com/@hi_caicai?source=rss-c2a2cc14885d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 23:51:01 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@hi_caicai/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[How Product Designers Can Move Faster in the AI Era]]></title>
            <link>https://medium.com/youmind-write-learn-studio/how-product-designers-can-move-faster-in-the-ai-era-0af2772f818f?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/0af2772f818f</guid>
            <category><![CDATA[product]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Fri, 05 Jun 2026 05:41:22 GMT</pubDate>
            <atom:updated>2026-06-05T05:41:22.126Z</atom:updated>
            <content:encoded><![CDATA[<p>Product designers are being left behind by a new development rhythm, and many people have not fully realized it yet.</p><p>Recently, YouMind went through a pretty fundamental shift in how we build. We started practicing a model where each module moves forward independently, with a few engineers owning it directly, instead of waiting for design files to be fully aligned before anything can move.</p><p>The result is that more modules can move at the same time, but quality and direction have also become harder to control. The role of the designer has become more passive. You want to keep up, but your original way of producing work is naturally one step slower than engineers who are now empowered by AI.</p><p>Instead of waiting to be left behind, it is worth asking how this role can move faster. These are some of my recent thoughts.</p><h3>Where we are now</h3><h4>The design process is not dead. It has moved upstream</h4><p>In my view, the traditional design process is not dead. What has changed is that the part designers used to spend most of their time on, drawing screens, polishing visuals, and communicating handoff details, has been compressed by 90 percent in the AI era.</p><p>The real center of gravity for designers has moved from execution to judgment.</p><p>There is less time left for drawing, but more time left for judgment, if you are able to handle it. So rather than saying designers are being eliminated, I would say the role is being pushed upstream. Either you move upstream with it, or you stay where you are and gradually lose relevance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YYZV_qp-Ys0U4p08m1OOww.png" /></figure><h3>What designers themselves will become</h3><h4>The identity of the designer is being redefined</h4><p>For designers, I think the future will move in two directions.</p><p>One type is the Design Engineer, a designer who can participate directly in product and engineering iteration through AI coding. Code becomes their design file. Between this person and engineers, there is almost no longer a clear concept of handoff.</p><p>The other type is the Design Strategist, a role closer to an architect. This person focuses on product direction, the personality and tone of AI models, and the final layer of quality control. They may not draw every screen by hand, but they decide what is worth building, what it should become, and when it is ready to ship.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qfSCUPm7A-2mD166pNvVdw.png" /></figure><p>The group that gets squeezed out is the group that only draws wireframes, only annotates UI, and does not engage with systems or models. This group still exists, but there is not much time left for them.</p><h4>Four traits that will become truly scarce in the AI era</h4><p>Looking forward, I think several types of designers will become especially valuable.</p><ul><li>People with extremely strong learning ability, who are smart, can absorb new information quickly, and can produce quickly.</li><li>People with strong aesthetic standards, who care about quality and can keep pushing quality forward.</li><li>People who can stay focused. The AI era creates endless temptations. Everyone wants to try everything. But the more chaotic things become, the more important it is to be able to sit down and do one thing well.</li><li>People with high judgment bandwidth. AI can generate infinitely fast, but human judgment bandwidth is limited. Whoever can keep making high quality decisions without burning out will go much further in the human in the loop process.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YYbAgCGX_gOwcqTCOkyjNA.png" /></figure><h4>AI does not have a point of view, but it will lower the passing line</h4><p>After talking about traits, it is worth talking about the designer’s moat.</p><p>What AI has not replaced so far is not just beauty. More precisely, it is point of view. AI is derivative. It remixes training data. It can become more and more visually polished, and tools like Nano Banana and GPT Image are already proving that. But it does not have its own worldview.</p><p>So the next layer designers should deepen has two parts. One is taste, knowing what is good and what is not good. That will still matter in the short term. The other is point of view, knowing why I choose this and not that. That will matter for a long time.</p><p>But on the other side, AI’s rapid progress in visual quality brings another risk that is not immediately obvious. It will make “good enough” design spread everywhere. In the past, if a company wanted to create a usable product interface, it at least needed someone who knew how to design UI and use Figma.</p><p>Now an engineer can use Claude Design or Stitch and generate something that looks acceptable in half an hour. The real danger is not that AI replaces great designers. The real danger is that it lowers the passing line for the entire industry. The distance between “good enough” and “actually good” keeps getting compressed by AI, which makes people with real craft harder to recognize.</p><p>In this environment, beauty and point of view will become more valuable, but only if you can make others see the difference. Designers in this stage not only need to make good work, but also need to explain why their work is different from AI’s default output. Translating the difference in quality into language other people can understand has itself become a new ability.</p><h3>How to run with teams and Agents</h3><h4>Running at the engineer’s pace is more reliable than asking engineers to wait for you</h4><p>In the AI era, trust in team collaboration needs to be rebuilt because the carrier of trust has changed. In the past, we trusted people. Is this person reliable? Can this person get things done?</p><p>Now we trust contracts. Is this design.<a href="http://design.md/">md</a> clear enough? Can this spec be understood by an Agent?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7nawROXI5nfvAwQFzuYBcg.png" /></figure><p>In this new collaboration model, AI amplifies the ability of solo players, but it also makes solo players more likely to drift away from the group. A wolf that loses the pack will eventually lose its fighting power. So designers need to be able to work independently, but also learn how to use the power of the team. Today, that team power is carried by contract files written for both humans and Agents.</p><p><strong>For designers, the most direct way to rebuild trust is not to ask engineers to “wait for design.” It is to enter the engineer’s rhythm yourself. Through vibe coding, designers need to make their speed, rhythm, and fidelity match the pace of engineering.</strong></p><p>Inside YouMind’s engineering team, this is already very common. Claude Code moves so fast that PMs and engineers may already be iterating on version three, while the designer is still sitting on version one of a high fidelity mockup. In that situation, trust is not built through meetings. It is built by whether you can keep showing up in the same rhythm.</p><p>I strongly agree with something Jordan Singer, who previously worked on product design at Figma, once said. Designers are seen as optional in this moment not because AI has replaced design, but because engineers are moving too fast and no longer have the chance to stop and listen to designers.</p><p>The real problem designers need to solve is not whether they will be replaced by AI. It is how to get into the same car as engineers, instead of being left behind at the station.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*U367bK_bBzP5ze7nfEQ64Q.png" /></figure><h4>What we design for has expanded from users to Agents</h4><p>Keeping up with the rhythm is only the first step. Going one step further, we need to realize that the object of design itself is changing.</p><p>In the AI era, I actually think documentation becomes more important, but its role has changed. It is no longer just a handoff artifact. It is the interface layer for collaboration. It is the only reliable interface between you, your team, and Agents.</p><p>This does not mean we need to write long documents to explain every product detail. It means we need to have very clear structural thinking and express it clearly. A good document is one where, after deleting 70 percent of the words, the remaining 30 percent still allows anyone to continue the work. Human teammates can fill in the gaps, and Agents can too. Whether that 30 percent is written by you or by AI does not really matter. The core is that the logic is clear and easy to understand.</p><p>Going further, the object of design is expanding from users to Agents. Beyond UX, there is now AX, Agent Experience. Can your interface be understood by an Agent? Can your documentation be called by an Agent? Can your API be composed by an Agent?</p><p>This is a completely new design capability. In the past, we were only responsible for humans. Now we are responsible for both humans and Agents. <a href="http://design.md/">design.md</a> is essentially the first contract of AX.</p><h3>Bringing it down to practice</h3><p>Since this article is about speed, here are a few things I think are worth advocating for if we want to move faster while still pursuing quality.</p><ul><li><strong>Treat </strong><a href="http://design.md/"><strong>design.md</strong></a><strong> as the new collaboration foundation.</strong> Build processes and standards that fit the AI era. Use files like <a href="http://design.md/">design.md</a> to constrain Agents, and gradually expand the old model of “communicating with people” into a new model of “collaborating with Agents.” I think this topic deserves its own article, because it is the foundation for everything else. Without it, none of the later actions can connect.</li><li><strong>The deliverable should move from UI to documents plus code.</strong> Documentation carries the logic, and it is also written for Agents. Code becomes the interactive form that designers can implement locally. If designers can fine tune something faster themselves, then engineers can choose to reproduce that implementation. But this depends on the scenario. When interaction is uncertain, code is useful. For visual systems and design systems that need long term consistency, Figma plus documentation is still the better deliverable.</li><li><strong>The time saved should go back into communication.</strong> The speed of the AI era does not mean we can be lazy, and it does not simply mean we should produce more. It means we can complete communication faster. AI production is infinitely fast, but human judgment bandwidth has not changed. So the time saved should be spent on more frequent syncing and sharing. Everyone needs to develop the instinct to keep others informed. Creating a Linear ticket does not equal collaboration. Real collaboration means making sure others know what changed in time.</li><li><strong>Do not trust AI blindly. You still need to understand the underlying logic yourself.</strong> I do not recommend the workflow where someone does not understand the principle at all and simply asks AI to generate everything. For example, when writing a YouMind product changelog, the worst workflow is to hand everything to AI and publish whatever it writes. That is not collaboration. That is outsourcing, and worse, it outsources the part you should be responsible for. The correct flow should be the opposite. First, you go through what changed in this version, why it changed, and what it means for users. Then you let AI help produce the first draft, and you revise from there. AI accelerates expression, not understanding. The understanding part can never be skipped. In the end, AI is an amplifier. It amplifies your judgment. If you have no judgment, what it amplifies is empty.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g2EHsHo60pgIyErN3KFPvQ.png" /></figure><p>Designers will not be replaced by AI, but they will be replaced by the version of themselves that cannot keep up.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0af2772f818f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/youmind-write-learn-studio/how-product-designers-can-move-faster-in-the-ai-era-0af2772f818f">How Product Designers Can Move Faster in the AI Era</a> was originally published in <a href="https://medium.com/youmind-write-learn-studio">YouMind Create &amp; Learn Studio</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Stop Hiring. Start Deploying Agents]]></title>
            <link>https://medium.com/youmind-write-learn-studio/stop-hiring-start-deploying-agents-9da591c84387?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/9da591c84387</guid>
            <category><![CDATA[openclaw]]></category>
            <category><![CDATA[agents]]></category>
            <category><![CDATA[teamwork]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Thu, 16 Apr 2026 06:46:30 GMT</pubDate>
            <atom:updated>2026-04-16T06:46:30.954Z</atom:updated>
            <content:encoded><![CDATA[<p>One evening, I was staring at my screen watching our designer work with Mindy on a design request, and something felt strange.</p><p>Mindy is our virtual agent. All we did that day was @ her in Linear and ask her to implement an updated design. A few minutes later, she had finished the changes, committed the code, merged it, deployed to staging, and pinged us to review.</p><p>The feeling wasn’t excitement. It was closer to vertigo. We’re a team of just over twenty people, but at that moment, several agents were running in parallel. Mindy was writing code. Pi was handling customer inquiries. Other agents were doing their own things. And us, the human colleagues, were in a meeting talking about next month’s product direction.</p><p>That was the first time it truly hit me: <strong>YouMind’s headcount is no longer just people.</strong></p><h3>From Single Tasks to Full Workflows</h3><p>From “help me do one thing” to “run this entire pipeline for me.”</p><p>I didn’t always think this way.</p><p>A year ago, we were still operating in the most traditional fashion: product, engineering, design, ops, support. Everyone had a clear role. AI was just a fancy assistant for us: draft some copy, generate a few images, write a bit of code. We used to joke about it, calling ourselves an AI native company while collaborating like it was still 2020.</p><p>The shift started with something mundane.</p><p>Around January, we were discussing how to get user feedback flowing to us more naturally, with proper context. The first instinct was to buy a CRM. But then I thought: why not build one? In about two days, using Claude Code as the foundation, I wired up email and a few other feedback channels into a dead simple AI native CRM.</p><p>It ran stable on my own setup for about a week, and it genuinely solved the email side of customer relationship management with almost no effort.</p><p>That experience changed how I saw things. The one person system is becoming real. Going from a need to a fully delivered workflow is no longer theoretical. I didn’t need anyone else involved, and yet it covered what used to require a small team.</p><h3>People Never Wanted Tools. They Wanted Outcomes.</h3><p>Once I realized I could replace a CRM in two days, I started rethinking a bigger question: if collaboration is changing, what happens to the tools we’ve always relied on?</p><p>The answer is brutal: <strong>when outcomes can be delivered directly, tools fade into the background.</strong></p><p>This isn’t about SaaS being bad. But think about it. Nobody ever wanted a tool. You buy a CRM because you want to close more deals. You use project management software because you want projects shipped on time.</p><p>What people want has always been outcomes. SaaS sells you the right to use a tool, charges you monthly per seat, and whether it actually delivers results is your problem. But when AI can deliver results directly, that whole logic falls apart.</p><p>Here’s what it feels like from my end: buying a system used to mean evaluation, procurement, deployment, training, weeks or even months of setup. Now we spin up an agent tailored to our own business in maybe two days. And it’s not some generic tool. It’s built specifically for how we work.</p><p>So I’m increasingly convinced that <strong>SaaS and apps will evolve into agents.</strong> The industry giants can see it. Most people just haven’t felt it yet.</p><p>Some will say: aren’t there already out of the box agent platforms? Honestly, my experience is that they feel more like toys. Fine for playing around, but if you need stable output in a real business, you still have to build it yourself. Which actually points to something interesting: building agents for others is going to be a real business.</p><h3>Building Agents Isn’t Hard. Everything Else Is.</h3><p>When it comes to building your own, most people assume it’s difficult. Our engineers were skeptical too. They figured anything vibe coded wouldn’t survive production. But once I took an open source agent framework and built our own internal platform with it, the team realized that solutions at this stage are a lot more accessible than expected.</p><p>That said, easy to build doesn’t mean smooth sailing. After hitting enough walls, I’d say there are three real challenges: <strong>correctness, runtime environment, and maintainability.</strong></p><p>Correctness is the one that hurts the most. A few days ago I was building a recording app. The AI kept insisting I needed to disable the sandbox to access audio data. It wasn’t until I sat down with our engineers that we found a solution that didn’t require turning off the sandbox at all, and it turned out Apple had their own demo for it. But the AI didn’t know that, and I didn’t either. When neither side knows, you can end up spinning your wheels in a dead end. You’ll probably solve it eventually, but not elegantly.</p><p>AI can do the work, but it doesn’t know what “right” looks like. That judgment call, at least for now, still belongs to humans.</p><p>As for runtime environment and maintainability: there’s no one size fits all answer for environments. File systems, terminals, and web APIs each have their place, and the key is figuring out what kind of workbench your agent actually needs. Maintainability is similar. Design the architecture from day one. Don’t wait until you’ve got a mountain of code before trying to fix things. And if you do need to start over, it’s only a couple of hours anyway.</p><h3>Permissions Make Agents Worse. Principles Make Them Better.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7sx6sRxZtOSPtq_IPtAxWg.png" /></figure><p>Here’s something we learned the hard way, and it might sound counterintuitive: <strong>don’t restrict your agent with permissions. Guide it with principles.</strong></p><p>My first instinct was the traditional approach. This it can do, that it can’t. Whitelists, blacklists. I was so worried about things going wrong that I layered on all kinds of guardrails. But what happened was the agent basically degraded into a glorified automation script. The whole point of an agent is that it can make judgment calls in ambiguous situations. If you eliminate all the ambiguity, what’s left?</p><p>So I changed my approach. Instead of telling it what it couldn’t do, I told it what to do when certain situations came up. Think of it like onboarding a new team member. You wouldn’t hand them a list of prohibitions. You’d say: when a customer complains, acknowledge their feelings first, log the issue, and if the amount is above a certain threshold, loop me in.</p><p><strong>Control becomes guidance.</strong></p><p>This design philosophy is exactly why Pi surprised us after launch. One time, we were about to send a follow up email to a user, and Pi jumped in: this user already received the same message on April 1st, sending it again would just be noise. Nobody programmed that rule. But guided by the principles we set, it made that call on its own.</p><p>This shift sounds simple, but it demands a lot from you. You need to understand your own business processes deeply: what the critical checkpoints are, where you can tolerate errors, and when a human absolutely needs to step in. A lot of people who say agents don’t work well are really saying they haven’t thought clearly about how their own business should run.</p><h3>Humans Own Direction. Agents Own Process.</h3><p>So when agents can do more and more, where does human value actually lie?</p><p>I’ve thought about this a lot. Skills are being leveled at an incredible pace. You can code, and so can an agent. You can design, and the agent can generate visuals. You can write, and it writes faster. If your value comes solely from being able to do a thing, you’re in a vulnerable spot.</p><p>But watching how our own team has evolved, I’ve noticed two things that remain irreplaceable: <strong>systems thinking</strong> and <strong>insight into user needs.</strong></p><p>Systems thinking means: can you architect a setup where multiple agents and humans collaborate without stepping on each other? Can you define clear boundaries and coordination patterns? This is the ability to make things happen. Agents don’t organize themselves.</p><p>Insight into user needs means: can you see what users haven’t articulated yet? AI is excellent at executing well defined tasks, but deciding what to build is still a human job. The more precisely you define the goal, the better the agent delivers.</p><p>So in the age of AI, what I value most is still <strong>the ability to deliver outcomes.</strong> The process gets dramatically compressed. That doesn’t mean process doesn’t matter. It just becomes more predictable. Agents handle execution. Humans define the goals, validate results, and adjust course. Process belongs to the agent. Direction belongs to the human.</p><h3>The Real Moat: Business Understanding Compounds</h3><p>One last thing I only recently figured out.</p><p>Pi has been running for about a month now, and I’ve noticed the quality of its responses has gotten noticeably better. Not because the model was upgraded, but because it has accumulated context about our business. It knows what customers ask most often. It knows what tone to use with different types of people. It remembers how much content each user has sent us, their story, their needs, even their offhand complaints. All of that has gradually become its memory.</p><p>A former colleague recently asked me where our “knowledge base” comes from. I told them it was built up organically through production. We didn’t start with pre written answers.</p><p>That made me realize something: <strong>the gap between an agent that’s been running with you for months and a freshly built generic agent is enormous.</strong> That gap isn’t about model capability. It’s about depth of understanding of your business.</p><p>I call it the “understanding moat.”</p><p>This might be the deepest moat in the agent era. Not a technology moat. Not a feature moat. Whoever understands your users and your business more deeply becomes irreplaceable. And this moat has a beautiful property: it only gets thicker over time. Technical advantages can be matched. Features can be copied. But an agent’s accumulated understanding of your business? Someone starting from scratch will need months to catch up.</p><h3>The Future Small Team: Don’t Hire People. Deploy Agents.</h3><p>Writing all this, I’m not sure every conclusion is right. Technology moves too fast. Half a year from now, some of these ideas might need revision.</p><p>But one thing I’m certain of: <strong>the “employees plus agents” model is no longer a question of whether for small teams. It’s a question of when.</strong></p><p>Mindy is still evolving. Sometimes her code isn’t elegant. Sometimes she drops the ball during iterations. Pi is still learning too. Every now and then it stumbles when facing complex customer emotions. But they get better every day.</p><p>And us, the human colleagues, we’re learning a new way of working too. Not doing everything ourselves, but defining direction, setting principles, and then trusting our agents to execute. That trust builds gradually. You start with a small task, observe, verify, feel confident, then hand over something bigger. It’s like onboarding a new hire. You wouldn’t give them your most important client on day one. But if they keep performing well, you give them more and more room.</p><p>The future small team won’t hire people. It will deploy agents.</p><p>Not because agents are better than humans. But because once you have agents, humans can finally do what only humans can do.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9da591c84387" width="1" height="1" alt=""><hr><p><a href="https://medium.com/youmind-write-learn-studio/stop-hiring-start-deploying-agents-9da591c84387">Stop Hiring. Start Deploying Agents</a> was originally published in <a href="https://medium.com/youmind-write-learn-studio">YouMind Create &amp; Learn Studio</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[YouMind: Finding Product Balance Through Subtraction]]></title>
            <link>https://medium.com/youmind-write-learn-studio/youmind-finding-product-balance-through-subtraction-8a4229de2875?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a4229de2875</guid>
            <category><![CDATA[tools]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[youmind]]></category>
            <category><![CDATA[product-design]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Tue, 27 Jan 2026 02:40:25 GMT</pubDate>
            <atom:updated>2026-01-27T02:40:25.951Z</atom:updated>
            <content:encoded><![CDATA[<p>After releasing the new product interface this week, I received two completely different types of feedback: some said it looked better and felt smoother, while others were upset asking “where did the features go?”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JZYapmQJsB0k9SgZuvH6bg.png" /></figure><p>As a product built from the ground up, YouMind has undergone three major interface changes in the past six months. I can’t say each one was perfect, but behind every change lies our vision for the product and incremental progress. So today, I’d like to take this opportunity to share some thoughts behind this latest redesign.</p><h3>The Courage to Subtract</h3><p>Sometimes we all say, “Man, adding a feature is so hard.” Sometimes we just can’t figure out what to build next. But for a product that’s already complex, the hardest part isn’t adding features, once users come on board, demand is usually abundant. At that point, <strong>the real test is what you’re willing to give up</strong>.</p><p>Many people don’t understand why we’d talk about cutting features when the product is doing fine and people are using it. There are many reasons, but for YouMind, which is still in the early stages of validating PMF, the thinking is quite simple: <strong>the earlier you are, the more courage you need to let go</strong>.</p><p>Because many features actually provide limited value but become obstacles on the path forward. They make the product bloated and rob it of the flexibility it needs. Imagine this: you quickly identify a potentially transformative need and want to evolve the product to address it, but you find that no matter how you try to change things, nothing works. That’s terrifying, because you haven’t left room for the product to adapt.</p><p><strong>A product’s flexibility comes from your courage to leave white space.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NhdrOUstcvDR2Oc36Q7Vyw.png" /></figure><p>Of course, giving up doesn’t mean mindlessly cutting features. Within a reasonable framework, it means <strong>letting everything return to its most natural state</strong>. Features often have logical relationships, and what we’re doing today is providing value to users at the right moment.</p><p>It’s like how everyone prefers to rename a file directly on the filename itself, rather than hunting for an “edit” button in some corner and then selecting “rename file” in a popup. That’s the most natural user behavior in the critical path.</p><h3>Why Such a Drastic Overhaul?</h3><p>So why did YouMind undergo such a drastic structural overhaul in version 0.8, even cutting many existing features?</p><p>Although we haven’t reached version 1.0 yet, YouMind was already facing some pretty serious problems:</p><ul><li><strong>Feature Bloat</strong>: We wanted too much, kept adding randomly, and ended up accumulating product and technical debt</li><li><strong>Too Many Entry Points</strong>: A single feature existed in locations A, B, and C, sometimes making it hard to tell them apart</li><li><strong>Complex and Impure Logic</strong>: Too many hidden relationships existed, making things less straightforward</li></ul><p>Because of these issues, we became hesitant in our actions. Many times, we should have handled requirements using the most natural logic, but instead we went in circles until even we got confused. No exaggeration — including myself, sometimes I wasn’t familiar with how certain features worked. This is just the reality of rapid iteration with parallel development.</p><p>While this is normal, it was definitely holding us back. If we wanted to quickly experiment with new features and scenarios today, this state made it hard to move fast. Especially in the fast paced AI wave, if we want to compete on speed, we sometimes have to slim down the product to ensure it stays in peak condition.</p><p><strong>Product growth isn’t about adding features, it’s about multiplying experiences.</strong></p><h3>The Detours We Took</h3><p>Actually, before this redesign, we also tried the simplest approach: “stuffing all features into the sidebar.” At the time, we thought that as long as we hid the features, the interface would become cleaner. But after repeated consideration, we found that even we couldn’t find the features we wanted. That’s when we realized: <strong>simplicity isn’t about hiding things, it’s about not needing them in the first place</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EIye6HG9-mU-sggGK_pD2Q.png" /></figure><p>Take our waterfall view feature, for example. It looked cool, and some team members really liked it. But in reality, many people just treated it as window dressing, purely aesthetic, not actually using it to manage content. Yet to maintain this feature, we had to ensure compatibility every time, and for a while we lost sight of priorities, adding a bunch of management features to what was essentially a display function.</p><p>So in this redesign, we quietly downplayed its presence. An old user commented: “Where did the waterfall view go? I loved that browsing experience!” I understand their disappointment and I’m aware of the consequences of this decision, but to keep the product from becoming overly complex, sometimes tough choices have to be made. Of course, the waterfall view is still there. It’s just become an auxiliary feature now, giving people a path to walk rather than making it the main highway.</p><h3>The Underestimated Competitive Edge</h3><p>So what’s left after we cut those redundant features? A purer, more beautiful product experience. And this change in experience even changed my own usage habits.</p><p>After this release, we spent some time observing feedback from users and team members. Most gave fairly positive feedback, saying the interface became more attractive and smoother to use. Of course, some people were upset because they couldn’t find previous features (which, as mentioned earlier, I had cut). Overall, the evaluation leaned positive.</p><p><strong>In today’s world of increasingly homogeneous product features, beauty and delightful experiences often become the key factors in user choice.</strong> When everyone can achieve “usable,” “easy to use” and “want to use” become the core differentiators. And this differentiation often comes from design restraint and attention to detail.</p><h3>My Creative Workflow Migrated to YouMind</h3><p>Speaking for myself, the tangible change is that I moved my creative workflow to YouMind.</p><p>I’m embarrassed to admit that I hadn’t really been doing text creation in YouMind before. I sometimes found it inconvenient to use, not because the features weren’t good. I believe our team’s work on AI writing is definitely superior to other products. The main reason was habit.</p><p>After the Yuque team disbanded years ago, I started migrating my document management to Notion. It’s been about two years now, and while Notion has its issues, I gradually developed a habit of creating text there. Previously, YouMind’s biggest use case for me was auxiliary reading and thinking. Because I’m the type who, while not a great writer, can just write whatever comes to mind, I didn’t really need a creation environment most of the time since that work was done in Notion.</p><p>But after this redesign, I’ve quietly started using YouMind to write documents I want to share. This isn’t because we’re the creators and I’m forcing myself to use the product. I believe many so-called product creators probably haven’t even touched 10% of their own product’s features.</p><p>But this time, both the interface and user experience genuinely moved me. You might say, “Come on, you’re just tooting your own horn!” Sure, I can’t rule out that possibility. But most importantly, <strong>I’ve started to believe that this version of our design, from my own perspective, has actually surpassed Notion’s experience</strong>.</p><p>My creative workflow migration happened naturally. This also validates from the side that this redesign genuinely moved me, and that feeling came from my most authentic inner response.</p><h3>Where Does This Satisfaction Come From?</h3><p>So if we look at this satisfaction from a design perspective, how do we articulate it? There are actually a few key points:</p><ul><li><strong>Extreme Simplicity and White Space</strong>: The current version of YouMind is very clean, with plenty of white space that lets you focus when creating or reading. For example, in the editor, we removed many of the originally complex toolbars, keeping only the most commonly used features, giving the entire interface more breathing room. When you open a blank page, you’re not distracted by a bunch of buttons and options, you can just start writing.</li><li><strong>Features Just Right</strong>: Not too many, not too few, basically appearing where you’d expect them. For instance, when reading materials, Pick (annotation) naturally appears when you select text, rather than being hidden in some corner. When you want to share content, the share button is right where you’d most easily find it. This “just right” quality makes the entire usage flow incredibly smooth.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qbzJNS9ugI3iIS4Wvt8zdw.png" /></figure><p>Of course, when it actually got to users, there was some feedback that differed from our design intentions. For example, some couldn’t find the waterfall view, others asked why Side Peek couldn’t open other content and only allowed materials. Behind this feedback are actually some of our thoughts on feature positioning.</p><h3>Feature Positioning Purity: The Side Peek Example</h3><p>Many times, it’s not that we couldn’t do it from a design perspective. It’s that I intentionally chose not to. I’ve debated this topic with the development team for a long time. Although everyone’s opinions sometimes differ, the basic assumptions are the same.</p><p><strong>We want to provide users with a purer product experience, which sometimes means not letting auxiliary features become more complex.</strong></p><p>Why doesn’t Side Peek have complex capabilities? Simple, since we defined it as Side Peek, it’s meant to be auxiliary in the moment, not a main feature. An auxiliary feature shouldn’t add to the overall complexity here, as that would make Side Peek steal the spotlight.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Eqf_cd-z7ziquDq65CGhPw.png" /></figure><p><strong>Good design isn’t about showing users all possibilities. It’s about showing them the one they need most.</strong></p><p>We didn’t design it to replace the main function, but rather to assist our learning and creation scenarios at the right moments in the main flow. Side Peek’s restraint is to keep the main function pure.</p><h3>In Closing</h3><p><strong>Product growth isn’t about piling up features. It’s about finding balance through subtraction</strong>. That’s the fundamental logic behind our redesign. Looking back at this entire redesign process, we also have some design insights to share:</p><ol><li>Features need clear boundaries: Main features should be just right, auxiliary features should be restrained, and shouldn’t steal the spotlight</li><li>Design should follow natural user behavior: Don’t make users adapt to the product, make the product adapt to users</li><li>Features require trade-offs: If a feature is only “potentially useful” rather than “definitely useful,” don’t build it</li></ol><p><strong>We believe that truly great products aren’t those with the most features, but those that understand users best.</strong> And this “understanding” often comes from polishing details and staying true to core values.</p><p>Of course, YouMind still has many issues, and there are many phenomena I can’t validate on my own in this document. But I hope that deep down, whether it’s me or the YouMind development team, we all maintain this restraint and vigilance, making YouMind a tool that is both elegant and just right.</p><p>If you’re also building products, I hope the thinking behind this redesign gives you some inspiration. If you’re a YouMind user, feel free to tell us where we can do better. After all, products are meant to serve users, and your feedback is what drives us forward.</p><p>Let’s work together to find product balance through subtraction.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a4229de2875" width="1" height="1" alt=""><hr><p><a href="https://medium.com/youmind-write-learn-studio/youmind-finding-product-balance-through-subtraction-8a4229de2875">YouMind: Finding Product Balance Through Subtraction</a> was originally published in <a href="https://medium.com/youmind-write-learn-studio">YouMind Create &amp; Learn Studio</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Hi Haye AI]]></title>
            <link>https://medium.com/@hi_caicai/hi-haye-ai-b3775f1a3c9e?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/b3775f1a3c9e</guid>
            <category><![CDATA[translate]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[tools]]></category>
            <category><![CDATA[macos]]></category>
            <category><![CDATA[assistant]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Wed, 20 Mar 2024 13:04:46 GMT</pubDate>
            <atom:updated>2024-03-21T01:20:22.430Z</atom:updated>
            <content:encoded><![CDATA[<p>Today, we are thrilled to introduce a new product to everyone: <strong>Haye, a system-level text intelligence assistant designed for macOS</strong>. By invoking Haye, you can streamline your reading, translation, and writing processes. Whether you are composing an important thesis or updating your Twitter, Haye is an indispensable assistant for you. <a href="https://haye.ai/">https://haye.ai/</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c80OuW2ERT9sOmqS1doLmA.png" /></figure><p>Now, let’s take a quick look at some of Haye’s features and the additional ones we’re developing:</p><ul><li>System-level toolbar: A toolbar that can be invoked anywhere, anytime. Once you complete the installation on your computer, you can summon it with the shortcut Command + E.</li><li>Supports any macOS application: Not just in Chrome, Haye can operate across any application on macOS. Imagine your Notes app suddenly having an intelligent toolbar like Notion AI, How exciting would that be?</li><li>Text content inquiry: Encountering unfamiliar text or want to delve deeper into a topic? Haye can leverage advanced models and search capabilities to help you explore further.</li><li>Text editing optimization: Whether it’s grammar or style, Haye can instantly enhance your writing quality, customizing content based on your intent and style.</li><li>Quick translation: Whether it’s a webpage, a PDF, or a name in your calendar, Haye simplifies translation, making language no longer a barrier.</li><li>Knowledge fragment management: Transform the information you inquire about and wish to learn into a knowledge base, and easily review it at any time. Haye also supports multi-platform synchronization, making knowledge management more effortless.</li><li>AI personalization: Haye is dedicated to adapting to each user’s unique style, accurately reflecting your personal characteristics, whether in information acquisition or writing preferences.</li><li>…</li></ul><h3>Origin</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*F30_PlpqKPSxQGCE5q5PIA.png" /></figure><p>There were quite a few turning points in the conception of this idea:</p><ul><li>We realized that the existing Chrome translation plugins were limited in functionality and lacked flexibility, which was unsatisfactory.</li><li>Inspiration came from everyday life, whether searching for an API at work or looking up the spelling of a word in daily life; we often found ourselves needing to make continuous inquiries.</li><li>Faced with Apple’s development team not yet integrating model capabilities into the macOS ecosystem, we thought it better to take action ourselves and provide sufficiently lightweight scenarios for users to enjoy the wonders brought by different large models on macOS.</li><li>We recognized that knowledge workers are early beneficiaries of large model technology. Although the products on the market are powerful, their usability is still not satisfactory. We wondered if we could integrate more deeply with the operating system, hence introducing the concept of the AI toolbar.</li><li>We hoped to find a way to automatically integrate daily fragmented knowledge for future use by personal models, which would be a significant advancement.</li></ul><h3>Perspective</h3><p>Of course, we’re not making a product just for the sake of making one; it reflects our key ideas and principles.</p><ul><li>As large model technology develops, its usage cost is gradually decreasing. Our goal is to accelerate this process, allowing more people to enjoy the convenience of large models at a low cost.</li><li>The entry barrier to the large model field is lowering. From our experience, most basic models are sufficient for daily work, so we are committed to integrating these capabilities more broadly into our daily lives and work.</li><li>The value of a tool lies not in its size but in its precision in solving problems. This philosophy also applies to the application of large models.</li><li>Even the most advanced large models can’t break through individual cognitive limits. We believe that breaking these limits and exploring unknown fields is a worthwhile pursuit.</li><li>Learning is a continuous process. We are constantly acquiring information and absorbing new knowledge every day to ensure that we can seize the best opportunities at the most appropriate time.</li></ul><h3>FAQ</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0UFOuFY_ApCy8g3HhdX87Q.png" /></figure><p>We anticipate there may be some questions, so we’re addressing them here in advance to help you better understand Haye</p><h4>Why not make a Chrome plugin, but a system-level tool?</h4><p>Although Chrome plugins are powerful, their scope is limited to the browser. Our vision is for Haye to transcend the browser and become a comprehensive tool seamlessly integrated with the operating system, providing a broader range of services.</p><h4>Why only macOS?</h4><p>Although Haye is developed using Electron, we currently only offer a macOS version. We strive for an experience close to native applications and are not content with merely copying and pasting code. After perfecting the macOS version, we will develop the Windows version with the same high standards.</p><h4>When is the full release expected?</h4><p>We need a bit more time. The overall framework is already operational, but some application experience details and large model adjustments will take up a considerable amount of our time.</p><h4>How can we experience it?</h4><p>We have already started a Waitlist. If you join the waiting list, we will send you the beta version to your email as soon as possible. You can also join our Discord to get our latest information.</p><h4>Is this a paid product?</h4><p>Yes, Haye is a paid product. We believe that a subscription model will ensure the software’s continuous development and establish a balanced relationship between users and developers. The final pricing will be determined based on feedback from the beta testing and cost considerations.</p><h3>Next Steps</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*r6CQRodvADBkULrxA4xKEw.png" /></figure><ul><li>We will start the beta testing this week. After the beta test, we will gradually receive our initial beta packages, although most features will still be in a non-functional state at that time.</li><li>We will launch a public beta in mid-April, by which time all features should be basically operational.</li><li>We plan to release the official version from the end of April to the beginning of May, adjusting based on our beta testing experience.</li><li>At the same time, we will introduce some new features, which will make text editing and reading even more exciting.</li></ul><p>Join the waitlist：<a href="https://haye.ai/">https://haye.ai/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b3775f1a3c9e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to use ChatGPT to build macOS applications without any code experience]]></title>
            <link>https://medium.com/@hi_caicai/how-to-use-chatgpt-to-build-macos-applications-without-any-code-experience-365ac048f0a0?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/365ac048f0a0</guid>
            <category><![CDATA[apps]]></category>
            <category><![CDATA[macos]]></category>
            <category><![CDATA[designer]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[text2code]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Tue, 30 Jan 2024 12:48:26 GMT</pubDate>
            <atom:updated>2024-02-20T09:51:29.635Z</atom:updated>
            <content:encoded><![CDATA[<p>Hello everyone, I am CaiCai, a product manager</p><p>After nearly a month of weekends delving into platforms and frameworks like Xcode, Electron, and Tauri, I finally created an app largely written by ChatGPT — astoundingly, over 95% of the code was generated by it. <strong>For someone without programming knowledge like me, this was not just a product challenge but a fascinating journey</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wHar92tv93KLsghbuQzGVw.png" /></figure><p>Although this app has a simple interface, it successfully demonstrates the potential of building applications with ChatGPT. However, ChatGPT can’t fully replace traditional development yet</p><p>It still lacks certain abilities, such as designing interfaces, so I designed and coded the UI myself. Additionally, it can’t handle peripheral project tasks, like packaging the app into a DMG file</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Df9aVeCqVFSAAWLFTobEAA.png" /></figure><p>Although this exploration has revealed many areas for improvement, it has also demonstrated the potential of using ChatGPT to build applications. Next, I will share how as a non-programmer, I used ChatGPT to build a macOS App and some of my experiences during the process</p><h3>Understanding the Basics</h3><p>Building a macOS App can seem quite complex, especially for those unfamiliar with programming. Before starting, it’s essential to understand some basic frameworks.<strong> For instance, which frameworks can you use to build a macOS App?</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zBPSyF-wWiiyU2LYHPtq1g.png" /></figure><p>Before I began learning, I had only heard of Xcode and Electron. Xcode is Apple’s official development platform for creating native apps but has a steep learning curve. Electron is essentially like a Chrome browser, making your app akin to a webpage. Later, I discovered Tauri — a smaller alternative, especially after finding Electron to be quite bulky</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jO168qvhsHFSeyqehkHpeA.png" /></figure><p>Understanding these is like unlocking levels in a game; you need to delve deeper to discover their secrets. Fortunately, ChatGPT can guide you in this, for example:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RLHlVms5T0tg05eqsonX3w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*466E6GW3X4WU0bBOZrhWXw.png" /></figure><p>Following ChatGPT’s guidance, you can gradually break down and complete the project. If you encounter problems, ask ChatGPT again for a solution. Compared to searching through numerous answers on Google, ChatGPT can guide you more methodically</p><p>Following ChatGPT’s guidance, you can gradually break down and complete the project. If you encounter problems, ask ChatGPT again for a solution. Compared to searching through numerous answers on Google, ChatGPT can guide you more methodically</p><ul><li><a href="https://www.electronforge.io/">Getting Started</a></li><li><a href="https://github.com/electron-vite/electron-vite-react">GitHub - electron-vite/electron-vite-react: :electron: Electron + Vite + React + Sass boilerplate.</a></li></ul><p>Both frameworks can quickly help you set up a basic structure. Then, you only need to understand and modify the code within (of course, you also need to inform ChatGPT about the framework you are using, and as much as possible, let it know the structure and code of the framework)</p><h3>Design Your Product Architecture, Clarify What You Want</h3><p>After mastering the basics, you can use ChatGPT to write your first App. <strong>However, as a product manager, I know that a good App doesn’t happen overnight. It requires clear goals and functions — this is your product architecture</strong></p><p>Of course, you can also describe naturally to let ChatGPT help you with this. If you are just experimenting, that’s fine, but if you want your App to eventually be practical, you still need to design it well</p><p>Take the Nustime App as an example. Its initial intention was simple: to calculate the remaining work time each day or week to help improve efficiency. On this basis, I added other features to enhance work focus</p><p>Break down the App’s features into small parts, like Nustime can be divided into:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EbDVURPyb-HEGtkD6PWAvA.png" /></figure><p>This method of decomposition allows you to provide ChatGPT with clear instructions. <strong>For example, ‘every day’s working time’ can be further divided into default time settings, remaining time calculations, and notification reminders</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Ahw8Zz-Beu2wX3EVmioBPw.png" /></figure><p>After breaking down the features, let ChatGPT handle simple tasks first, based on your understanding of complexity. For example, the settings and notification functions in the list are more complex than pure calculation functions and are not suitable as initial commands. If you are unsure about the complexity of a feature, you can consult ChatGPT, which can provide guidance.</p><p>For instance, if you don’t know how to start with the initial framework, you can paste your requirements and the framework’s code into ChatGPT, and let it tell you how to solve it. Below is the implementation process for a requirement to display local time:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mqR2OpYnKBOnl7S1PXNEjQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TebiqJho69cGTTBcricXuQ.png" /></figure><p><strong>The key is to start with simple features, gradually introduce your ideas to ChatGPT, and continue to refine them on this basis. Only then can ChatGPT better understand and implement your needs</strong></p><p>For example, as the first feature of Nustime, you can eventually summarize a simple command:</p><p><em>“I want to design a function for working hours, with the default time set as 9:00–18:00. Then you will get the current time, calculate the remaining time of each day, and output it in the format of 00:00:00. Please provide the complete code and explain the steps to implement it”</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bsZjwcw0ry5kvnywEypcbA.png" /></figure><p>When you paste the code generated by ChatGPT into the code editor, you can run the corresponding effect. The following image shows the Remaining Time generated by ChatGPT, with 02:18:01 as the calculated result</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vp4w_T-Hejd8X_KAcBAQnw.png" /></figure><h3>First build the Interface, Then the Functionality</h3><p>In my experience using ChatGPT, I found that it has limitations in understanding user interfaces (UI). Therefore, I suggest first conceptualizing the product’s basic framework, which will help you provide clearer instructions to ChatGPT. My practice has taught me that it’s best to first sketch out the interface and then request corresponding functionality based on it. This approach can help ChatGPT better understand your needs</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4LS10De7uC7q61kfSkNsYA.png" /></figure><p>Although ChatGPT can understand basic layouts if you input a design image, it still lacks in front-end details. This means you may need to spend time adjusting and perfecting these details. The following image shows the initial interface framework</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LTyKvnsIKTUPoNTdoO9CNw.png" /></figure><p>If your app doesn’t involve complex interface design, like just a Menu Bar menu function, the whole process becomes much simpler, and you only need to focus on implementing logic. My first attempt using Xcode was like this, where I created a simple application without any interface design</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TU6RgPDT2t2-fAme03mIBg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*27mcSCPfH5w5S5Q-accWEQ.png" /></figure><p>After multiple trials, I found that projects with clear UI designs are easier for ChatGPT to understand, significantly reducing error rates and enhancing our work efficiency</p><h3>You’re Not a Trainer, but a Requester</h3><p>When using ChatGPT, I often feel like I’m training a pet, trying to teach it to walk or talk. But such efforts are futile. ChatGPT is just a large language model filled with knowledge, and its role is merely to utilize this knowledge to converse with you</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/925/1*AmKwQMleCru1KHpqNdgHTg.png" /></figure><p>I once mistakenly thought that improving my way of speaking and adding polite phrases would make it more intelligent. But in reality, it made little difference. ChatGPT’s responses are still based on its programming</p><p>For instance, I once tried to guide it to help me change the Menu Bar icon in an Electron App, but it told me that Electron does not support this feature. However, when I checked Electron’s official documentation, I found that it was feasible through Tray Icon. At that moment, I realized that I needed to tell ChatGPT the correct approach for it to give me the answer I needed</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yb6AdL5XQJXZg4RbmfOgjA.png" /></figure><p>For most beginners, this might be a challenge because they might not be familiar with the technical ecosystem. As a novice, I usually try various solutions and keep asking ChatGPT if there’s a better way</p><p>At the same time, I also consider the timeliness of the technology dependencies. If a component hasn’t been updated in five years, I generally insist on avoiding it, thus changing ChatGPT’s recommendation strategy</p><p>Remember, ChatGPT’s data is not always up-to-date. Sometimes you need to look up the latest version information and then ask ChatGPT if it can provide the corresponding answer</p><p>In summary, when using ChatGPT, you’re not a trainer but a requester. You need to be clear about your requirements, rather than completely relying on ChatGPT’s answers. In many cases, you still need to do additional research and verification on your own</p><h3>Check Logs Frequently, Learn to debug</h3><p>For beginners, facing a series of errors when running code in the terminal can be the most frustrating thing. Often, our understanding of errors is very limited, and even if we find seemingly similar solutions on the Internet, they may not necessarily apply to your specific situation</p><p>Luckily, with ChatGPT, a powerful assistant, solving these problems becomes much easier. Combining the features of <strong>Visual Studio Code + Github Copilot</strong>, I found it quicker to solve these coding puzzles than before. You just need to select a specific paragraph or the entire code and paste the error message into the chat window, and ChatGPT will quickly generate a solution based on your code. I’ve solved issues by repeatedly trying in this way until I found and fixed the problem</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kr52aR-tIL1AcuZNKeuVzA.png" /></figure><p>But merely looking at error messages is not enough. Some errors may involve whether the program is actually running. For example, if I add a settings feature to the app, allowing users to customize their working hours (e.g., change from the default 09:00–18:00), it becomes crucial to add log output. After implementing the settings feature, I would try operating it, and if I saw the corresponding time log output in the terminal, it would mean I had successfully implemented this feature</p><p>I’m not sure if there’s a more efficient method for experienced programmers, but for beginners like me, this method is sufficiently effective, helping me quickly locate and solve problems</p><h3>Don’t Be Discouraged; Victory is Just Ahead</h3><p>I understand that it’s natural to feel frustrated when using ChatGPT, especially after repeated attempts that often fall short of expectations. Particularly for beginners unfamiliar with programming, you might hope that ChatGPT could output a complete set of code, allowing you to easily copy and paste to complete the task. However, due to output length limitations, ChatGPT sometimes might provide only a basic implementation scheme. While this might be enough for experienced programmers, it’s like a tug-of-war for beginners</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9p6WYcSIDSumCGP770itQw.png" /></figure><p>After a long time of trying and failing, you may become irritable and increasingly dissatisfied with ChatGPT’s output. But at such times, I tell myself not to lose heart. Perhaps it’s time to think differently. If ChatGPT can’t output complete code, then let it clearly tell you which lines of code have been modified. This way, you only need to do a few more input and copy operations</p><p>Also, if you find that you’re consistently not achieving satisfactory results, don’t hesitate to try different solutions with ChatGPT, such as changing the implementation logic or outputting in segments</p><p>There’s a Chinese saying, “Dripping water wears away a stone,” which aptly describes the patience and persistence required when dealing with a not-so-smart ChatGPT. If one approach doesn’t work, then try another. After all, persistent efforts will eventually lead to victory</p><h3>Conclusion</h3><p>After these shares, if you’re also inspired to use ChatGPT to complete your own projects, the following tips might be helpful. I wish you success:</p><ul><li><strong>Start small: </strong>Don’t conceive a grand product; that’s beyond ChatGPT’s capabilities. Start with simple functionalities, like checking the date or making a countdown</li><li><strong>Limited knowledge updates: </strong>Remember, all large language models have outdated information. Don’t expect them to provide the most timely knowledge; this part still requires your own research and supplementation</li><li><strong>Build a basic framework: </strong>No matter what you do, a foundational framework is needed, and product development is no exception. We need to delve deeper and keep describing your product concept to ChatGPT</li><li><strong>The right way to cooperate: </strong>There are no “dumb” people or completely “foolish” AI models. The key is to find the interaction method that suits you best</li><li><strong>ChatGPT is just an assistant:</strong> Don’t regard ChatGPT as an all-powerful solution. Currently, it’s just a helper; the real decision-maker is you</li></ul><p>Tools used in this article:</p><ul><li>ChatGPT</li><li>Github Copilot</li><li>Visual Studio Code</li><li>Xcode</li></ul><p>Nustime is currently in the final stages of development, such as packaging and signing, and will soon meet everyone. Additionally, we have launched another application, EasyDevo, but unfortunately, due to the complexity of this application exceeding expectations, ChatGPT’s involvement in this project was limited</p><p><a href="https://github.com/boring-design/nustime-distro/releases/tag/v0.1.5">https://github.com/boring-design/nustime-distro/releases/tag/v0.1.5</a></p><p>We invite everyone to try EasyDevo, and we will soon launch several new applications</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mxifFFI0yrklHLh_9piJIg.png" /></figure><p><a href="https://easydevo.boringboring.design/?utm_source=medium&amp;utm_medium=post">https://easydevo.boringboring.design</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=365ac048f0a0" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[EasyDevo: The Developer Tool You Might Need]]></title>
            <link>https://medium.com/@hi_caicai/easydevo-the-developer-tool-you-might-need-a70dfd6ee3e1?source=rss-c2a2cc14885d------2</link>
            <guid isPermaLink="false">https://medium.com/p/a70dfd6ee3e1</guid>
            <category><![CDATA[tools]]></category>
            <category><![CDATA[dev]]></category>
            <category><![CDATA[macos]]></category>
            <category><![CDATA[apple]]></category>
            <category><![CDATA[apps]]></category>
            <dc:creator><![CDATA[CaiCai]]></dc:creator>
            <pubDate>Sun, 28 Jan 2024 10:38:20 GMT</pubDate>
            <atom:updated>2024-02-20T09:52:52.691Z</atom:updated>
            <content:encoded><![CDATA[<p><a href="https://easydevo.boringboring.design/?utm_source=medium&amp;utm_medium=post">https://easydevo.boringboring.design/</a></p><p>We’re thrilled to launch our first app, crafted specifically for developers. This handy tool addresses the growing storage challenges in your development environment, easing the burden of project dependencies, SDKs, and older Xcode versions. It doesn’t stop there; integrated with system monitoring and project analytics, EasyDevo helps you better understand your device performance and review your project contributions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mxifFFI0yrklHLh_9piJIg.png" /></figure><p>Let’s dive into how it works:</p><h3>Junk Cleanup</h3><p>Unlike traditional file cleaners, EasyDevo focuses on tidying up your development space, including project dependencies, SDKs, and some old files on Xcode. We’re committed to constantly refining this process, freeing up storage to rejuvenate your development experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CEDECfSjcOdRB71y7O5INg.png" /></figure><p>Usage: Select the items you want to clean in Dev Tools and wait for the scan. EasyDevo will request file access permission on first use, which requires your authorization.</p><p>Currently supports cleaning:</p><ul><li>Project dependencies and build outputs: Node.js node_modules folders, Rust target folders.</li><li>Xcode: Device Support, Simulators, Derived Data, Archives, and Caches.</li><li>SDK: Integrated rustup and nvm for cleaning Rust toolchain and Node.js.</li></ul><h3>System Monitoring</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HdC_NSa8bbqwYMRY4RnhIQ.png" /></figure><p>Our lightweight system monitoring tool effortlessly keeps you informed about key performance indicators of your computer, including network, CPU, RAM, and battery health. Key features include:</p><ul><li><strong>Network:</strong> Monitors total download/upload, real-time speed, displays Local/Public IP, and Proxy IP (if enabled).</li><li><strong>Project Storage Summary:</strong> Shows the storage proportion used by your projects and the main programming languages used.</li><li><strong>CPU &amp; RAM:</strong> Real-time monitoring of CPU usage, temperature, and RAM usage percentage, aiding in efficient system core performance monitoring.</li><li><strong>Battery Health:</strong> Provides maximum battery capacity, overall health, cycle count, and estimates battery wear based on charging history.</li></ul><h3>Project Analysis</h3><p>Beyond cleaning and system monitoring, EasyDevo includes a project analysis module. This feature aims to provide engineers with an in-depth review of their projects, showcasing individual contributions and time invested. Easily identify which files have seen the most changes recently.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ElAs6nRehVNNTq-6DtAw9g.png" /></figure><p>Furthermore, our project analysis allows you to view the time invested, storage usage, and language analysis for each local project. It also displays recent updates and the active status of projects.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RGhdqcJcHWLnyL5RfB69Lg.png" /></figure><p>Our goal with this module is to help every developer gain clear insight into their work contributions and outcomes.</p><h3>What’s Next</h3><p>As our inaugural software, EasyDevo will continue to evolve, currently focusing on enhancing system analysis, junk file cleanup, and project analytics. We’re also about to introduce a Menu Bar feature for quick access to essential information. In the future, we plan to launch versions for different operating systems to cater to various development environments.</p><p><a href="https://easydevo.boringboring.design/?utm_source=medium&amp;utm_medium=post">https://easydevo.boringboring.design</a></p><p>We welcome and look forward to your valuable suggestions. Feel free to contact us or join the product discussion at: <a href="https://discord.gg/f2sTUD7HvH">https://discord.gg/f2sTUD7HvH</a></p><p>We also need your support</p><p><a href="https://www.producthunt.com/posts/easydevo?utm_source=badge-featured&amp;utm_medium=badge&amp;utm_souce=badge-easydevo">https://www.producthunt.com/posts/easydevo</a></p><p>Boring Design Studio</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a70dfd6ee3e1" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>