Lakshay Nagpal

I've always had things to say but never posted. Not because I didn't care but building a startup always felt more urgent than writing. But a founder without a voice is invisible. So I built a bot that pings me twice a day and won't let me forget. Systems > Willpower

Back to badminton after a year of pickleball. Took a few days to stop playing like a confused hybrid athlete. Now I’m back to the level where I left. Time to actually improve the shots now.

Functional design > aesthetic design. Aesthetic design makes the product look better. Functional design makes the product make sense. I felt this today while talking to a few students about Homi. I was excited to tell them: “Homi now checks your Maths answer and speaks the feedback in audio.” I thought that would be the wow moment. It wasn’t. Then I said: “It also tells you what question you should practice next.” That clicked immediately. Because the student was not thinking: “Will this product speak?” They were thinking: “What should I do now?” That was the product lesson. As builders, we fall in love with the feature that was hardest or most exciting to build. Users fall in love with the thing that removes their next confusion. For us, audio was the innovation. For them, the next question was the relief. Good aesthetics create attention. Good functionality creates clarity. The best products do both. But if the product does not make the next step obvious, beauty won’t save it.

Homi now won’t JUST check your answer, but it will: 📝 Give personalised feedback ✅ Suggest you questions to practice next Just like a teacher would do 👩‍🏫 🔗 Try now - LINK IN BIO!

Try here: play.google.com/store/apps/det…

In most products, audio is decoration. A click sound. A success chime. A notification tone. In education, audio can be much deeper. When a student gets feedback, the job is not only to show what went wrong. The job is to bring their attention back. That is why we added audio feedback in @HeyHomiApp. After checking a Maths answer, Homi now speaks to the student and tells them what to practice next. Not just: “Here is your mistake.” But: “Here is what you should do now.” That small shift matters. Because learning does not end at feedback. It starts after it.

@gregisenberg Thanks for this @gregisenberg Saving it and sharing with the team to get started on it now.

@AndrewYNg Prompting is becoming less about “getting a better AI answer.” It is becoming a way to think better. Breaking problems down, giving context, creating checks, asking for plans - these habits improve how humans work with teams too.

How we prompt AI is very different in 2026 than 2022 when ChatGPT came out. I'm teaching a new course, AI Prompting for Everyone, to help you become an AI power user — whatever your current skill level. It covers skills that apply across ChatGPT, Gemini, Claude, and other AI tools. How to use deep research mode for well-researched reports on complex questions. How to give AI the right context, including more documents and images than most people realize you can provide. When to ask AI to think hard for several minutes on important decisions like what car to buy, what to study, or what job to take. And how to use AI to generate images, analyze data, and build simple games and websites. I also cover intuitions about how these models work under the hood, so you know when to trust an answer and when not to. Along the way, you'll see flying squirrels, a creativity test, some of my old family photos, and fireworks. Join me at deeplearning.ai/courses/ai-pro…

We are going deeper into education every day. Not because it is an “AI opportunity.” But because we are all byproducts of education.. Uniforms. Morning assembly. Maths homework. Board exam fear. For me, it is even more personal. I come from a family of teachers. I grew up hearing the teacher’s side too. The strictness. The patience. The repetition. The joy when a student finally understands. That is the feeling we are trying to build with @HeyHomiApp. A student solves a Maths question in their notebook. Homi checks it with feedback on the actual answer, speaks the feedback in audio, and tells them what to practice next. Evaluation → Attention → Next practice. We released the new version today. Try it here: play.google.com/store/apps/det…

@thisisgrantlee Things have always changed. The difference now is time compression. What earlier took decades now takes months because the world can produce, test and ship more in the same amount of time. AI didn’t invent change. It compressed the timeline.

@lennysan @Nerdi_Yogi Creative builders with Product Sense ✅ Product Managers ❌

Claude Code eng leader @Nerdi_Yogi on the two profiles she's hiring for now: 1. Creative builders with product sense 2. Deep systems experts (for the hard parts)

@DanielMiessler Most companies don’t have an AI problem. They have a context problem. Workflows live in people’s heads. Nothing is documented. No one understands the full loop end to end. If AI can’t understand the system, it can’t improve the system.

@thesangwookwon Interesting product @thesangwookwon Let me give it a try

Absolutely Lakshay! For me, giving full, comprehensive context was the real bottleneck. That's why I built Sherlock: The AI Note that knows your context. It learns from your screen, writes automatically, and answers your questions with an understanding of your context. Try Sherlock Now. getsherlock.xyz

AI didn’t just make me better at prompting. It made me better at working with people.

AI didn’t only improve my output. It improved my operating system. Better questions. Clearer thinking. Sharper plans. More checks. Prompting is not just an AI skill. It is slowly becoming a management skill.

When I discuss with my team now, I’m much more conscious of the flow: What is the goal? What are the constraints? What are the assumptions? Where can this fail? What checkpoints do we need?

@sflorimm Features get copied in weeks. What doesn't: speed of feedback loop, taste, supply chain, data, offline connections, & grit to stay when vibe coders leave. The fundamentals haven't changed. Same thing that killed Kodak or Blackberry, just compressed from years to months now.

What makes a startup uncopyable?

Speed and quality aren't tradeoffs anymore. They're both table stakes. The new edge? Shipping fast AND having taste. Building quickly AND knowing what's worth building.

The Hoffman quote almost inverts in the post-AI world. Speed still matters. But speed alone looks like a vibe-coded product now. The "basic" quality bar has gone up. What used to impress as a fast launch now just looks unfinished.

Reid Hoffman said "If you're not embarrassed by the first version of your product, you've launched too late." That was true. In a world where building was hard, shipping fast was the edge.