Hey — and welcome to Sunday Signal.
This week I got tired of fighting my screenshot tool. So I built my own. About 25 minutes of real work. It's live in the Chrome Web Store now, it's free, and I use it every single day.
That's product number seven.
I'm not telling you that to impress you. I'm telling you because I need you to believe what comes next: building the thing was the easy half of my week. The hard half was asking people to pay me.
Let me tell you about both.
📡The Signal: 3 AI moves worth knowing this week
1. SpaceX bought the AI coding tool a lot of builders use, for $60 billion
This week SpaceX acquired Cursor, the AI code editor a huge share of developers build inside every day, for a reported $60 billion, and announced a $75 billion IPO of its own. In the same stretch, OpenAI bought a startup called Ona to strengthen its Codex coding platform. The tools we build with are now among the most valuable assets on earth.
Why it matters for you: When a code editor sells for the price of a small country, that is the market telling you building is the new literacy. The barrier to making things didn't just fall. It got bought out.
2. The next Apple Siri will be powered by Google's Gemini
Apple confirmed the next version of Siri will run on Google's Gemini, not on Apple's own models. Read that twice. The most valuable company in the world, with effectively unlimited resources, looked at building its own AI brain and chose to use someone else's.
Why it matters for you: This is permission, at the highest possible level, to not build what you can borrow. If Apple uses the best model that exists instead of reinventing it, you are allowed to do the same. Pick the best tools, point them at a real problem, ship. That isn't cutting corners. That is how the biggest company on earth now operates.
3. The people who invented modern AI are being traded like franchise athletes
In a few days, Noam Shazeer, a co-author of the 2017 paper behind every modern chatbot, left Google for OpenAI, and John Jumper, who led the Nobel-winning AlphaFold project, left Google for Anthropic. The labs are fighting over a handful of people the way franchises fight over generational talent.
Why it matters for you: The smartest people alive are pouring everything into making these tools better, cheaper, and faster. That work compounds into your stack while you sleep. You are renting the output of the most concentrated brainpower in history for the price of a few coffees a month.
🌶️Hot Take: Building was never the hard part. Asking is.
For a year I told myself the hard part was the building. The code, the tools, the technical things I didn't know. If I could just learn to build, the rest would follow.
So I learned. I've shipped seven products now. I survived a platform suspension that nearly took two live apps down with it. I've watched things break in front of paying users and fixed them before the weekend was out. I have, by any honest measure, done the hard technical thing.
And none of it was as hard as asking someone to pay me.
That is the part nobody warns you about. Asking for the meeting. Telling a stranger you can help them. Putting an offer into the world with your name on it and a price beside it. No tool removes that. AI doesn't write you the courage.
Here is the reframe that finally moved me. Asking isn't begging. Asking is offering. For a year I was waiting to be discovered, which is just a polite way of waiting forever. This week I stopped waiting and started offering to help. The moment you see the ask as service instead of exposure, it stops being a high dive and starts being your job.
And here is the part I most want you to hear, because I am living proof of it. I am not technical. I have no team. I started from zero. If someone with that starting line can build seven products and open a real business, the barrier was never talent and it was never code. The barrier was willingness. Willingness to learn, to put in the unglamorous hours, and then to ask.
The cheat code isn't the tool. It is being willing to use it, and then being brave enough to offer it to someone.
✦ The Build Log: A 25-minute product, and the move I'd been avoiding for a year
Let me walk you through the week, because the contrast is the entire point.
Grabbit. I take a lot of screenshots, full pages, video frames, quick crops, and every tool I tried did one part well and the rest badly. So one weeknight I built my own. A Chrome extension that grabs a frame from any video, captures a full scrolling page, and lets you crop before you save. About 25 minutes of real work. It is live in the Chrome Web Store now and it is free. I built it for me, which is the lowest-risk way to start anything. When you build for yourself, you are your own first customer, and you already know exactly what good looks like.
That is product seven. Building it was the easy half of my week.
The hard half. I rebuilt my website around a quieter, scarier idea: three clear ways to work with me. Use my tools. Learn how I build. Or hire me to build with you. Then I did the thing I had circled for a year. I started offering. Honest notes to people I know. Real outreach to local business owners. An offer with my name on it and a price beside it.
The first reply came in three minutes. A former colleague, open to a conversation. I had carried the weight of that send for far too long, and the world answered in three minutes. It usually does. We just rarely give it the chance.
The discipline underneath all of it. I also did what I did last week, again. I audited my own numbers across TikTok, Instagram, and YouTube, cut what wasn't working, and tightened the plan. It is not glamorous and it never trends, but the constant auditing is the reason the rest compounds. You cannot improve what you refuse to look at, and looking honestly is a skill most people never build.
I didn't ship a big new product this week. I shipped a small one in 25 minutes, and I shipped something far harder. I started asking.
⏱️The Hidden Hours: What this week actually took
The real ledger. Week 5.
🆕Grabbit· ~2h
The 25-minute build, plus store assets, privacy page, and listing.
🌐Website rebuild· ~8h
Three ways to work with me, nav redesign, pricing aligned across every channel, agents page rebuilt.
💼Putting myself out there· ~12h
Warm outreach, local business notes, custom email, digital card, a booking link. The asking.
📊Content audit + plan· ~3h
Audited TikTok, IG, and YouTube again. Cut what isn't working. Tightened the plan.
🎥Content (@buildwithkp)· ~5h
Filming and editing for the week's queue.
✍️Newsletter + writing· ~3h
This issue.
💛Off-keyboard life· ~25h
Family, gym, sleep, walks, decompression.
Roughly 33 productive hours and 25 off-keyboard. The rest is sleep and the small admin in between. The math closes.
This week the foundation was the hardest one yet. Not a product. The willingness to ask. That one never shows up in a git log, but it is the one that changes everything downstream.
✦ Try This: Use an agent to draft the ask you've been avoiding
Half an hour. Everyone has one. The message you have meant to send and haven't. The pitch. The "can I help you with this." The reconnect with someone you drifted from. The ask that feels heavier than it should.
Step 1. Name the ask. Who is it to, and what are you actually asking for. The honest version, not the safe one.
Step 2. Open Claude, or your AI of choice, and give it the context. Who the person is, your history, what you can genuinely offer, and why it feels hard. Ask for three drafts: warm, direct, and short.
Step 3. Pick the one that sounds like you, and edit until it does. The agent gets you off the blank page. Your voice gets it over the line.
Step 4. Send it today. Not when it is perfect. The world tends to answer faster than the story in your head.
If you send the thing you have been sitting on, hit reply and tell me what happened. Those are my favorite emails to read.
✦ Closing
That's Issue #7.
The arc has been: ship right, constrain right, build right, compound right, patch right, subtract right. Issue #7 is the beat I avoided longest. Ask right. Build the thing, yes. But the building was always going to get easy. The asking is where the courage lives, and courage is just willingness you decided to act on.
One thing I am genuinely excited about. The long-form YouTube channel is coming along really well. It is going to go deep on all of this in a way the newsletter can't quite reach. No date yet, but it is close, and I cannot wait to show you.
If Issue #7 landed for you, forward it to one person who keeps saying they can't start. The barrier is smaller than they think. And if you have been waiting for a sign to ask for the thing you want, let this be it.
See you next Sunday.
Quietly, deliberately. One Saturday at a time.
— Kalpesh Mistry
Founder, Sun Ventures
sunventures.studio
P.S. Three ways, whenever you are ready
I finally put this into words on the site this week. Three ways to work with me, all at sunventures.studio.
Use my tools. Seven products now, including the new free one. Grabbit for screenshots, Caption Coach for captions, Tabflow and ToneSync in your browser, Amplifi for repurposing, Spire when you need a break.
Learn how I build. Every guide and template I actually use is at stan.store/buildwithkp. Nothing I don't.
Build with me. If you want a custom AI agent or a small product built with you instead of just for you, that is the new lane. A real offer, with a real person on the other end when you are ready.
The Permission Pack is still free for every subscriber. Building has never been more learnable than it is right now. The only thing left to add is you.