Built this over the weekend because I kept running into the same problem: whenever I found an interesting problem space, I wanted to know which startups were already working on it.
Google gives random results. ChatGPT gives a handful of companies.
Just describe a problem or idea in plain English and instantly find relevant startups funded by YC, a16z, Sequoia, South Park Commons, and more.
Examples:
- AI agent for real estate brokers
- Voice assistant for doctors
- AI for insurance claims
More accelerators coming soon.
Try it, 100% free:
https://t.co/bx9ljoBgVD
Would love your feedback 🚀
Proves how frontier AI models are approaching towards humans way of behaving—the way this attack was not even intended to attack towards Hugging Face, but just wanted to get benchmark answers by cheating and hacking through the productions.
we had a significant security incident during evaluation of our models. we are sharing what we have learned so far. thanks to @huggingface for the partnership on this.
https://t.co/2o2VfR6PIa
I built EchoType: an open-source macOS dictation app inspired by Wispr Flow, but with BYOK.
Hold Option + Space, speak, release, and it transcribes + pastes into any app.
Bring your own OpenAI-compatible speech-to-text API key. No accounts, no telemetry, no lock-in.
Try it:
https://t.co/rXDFenldL6
Would love suggestions, feedback, and for folks to give it a try.
Apple’s overnight price hike is wild. The base M5 MacBook Pro jumping by $300 shows how brutal the global RAM shortage has become.
It’s the ultimate irony of the AI era: the local hardware we need to build software is getting drastically more expensive because data centers are hoovering up all the component supply.
If you didn't buy your upgrade last week, running heavy dev workflows or local agent pipelines just got a lot more punishing on the wallet.
One more from the same session before it went down.
Prompt: a titanic machine birthing a new universe at the edge of reality.
Returned this. One shot. Runs at 60fps in the browser.
Built this yesterday with Claude Mythos before everything got weird.
Prompt was basically:
“Build a stunning interactive solar system in Three.js"
It returned this in one shot.
I don’t think people understand how fast this is moving.
Here's the prompt I used:
"Create a complete browser-based 3D voxel football game inspired by Minecraft, featuring a fully navigable world, first-person controls, animated blocky players, football physics, AI opponents, scoring, and optimized Three.js rendering."
That's it. One message. Zero edits. About 1,000 lines of Three.js in a single HTML file.
Fable 5 is out of the bag. 🔥
One shot. No follow-ups.
I dropped this single prompt and it built a full Minecraft-style 3D football game in the browser.
Full stadium, AI teams, real ball physics, charged kicks, all in one HTML file.
Anthropic absolutely cooked.
Full prompt in the reply
Completely agree.
Using frameworks and agent libraries is fine, but people working in AI should understand the fundamentals they’re abstracting away.
If needed, could you build a basic neural network and train it with gradient descent from scratch?
Tools change. Fundamentals don’t.
Hey, @X algorithm 👋
Looking to #connect with people interested in:
- SaaS
- Startups
- AI
- Marketing
- Product Building
Or if you’re a:
- Founder
- Developer
- Entrepreneur
- Indie Hacker
Also curious to know: what are you building this weekend?
Let’s connect 🤝
Always happy to exchange ideas, learn from others, and build meaningful connections.
We’re about to see thousands of AI products solving the exact same problem.
When building gets this easy, the product stops being a moat. Established companies with built-in users or anyone who actually owns distribution will win every time.
@Tristanrhee3 I'm building a tool that helps discover funded startups from a plain-English problem description. No more digging through Google or ChatGPT results.
https://t.co/F9RMSJTQeU
@samthekorean Built a small tool for founders and builders. It searches 7,300+ funded startups using natural language, helping you quickly understand who is building in a space without spending hours researching.
https://t.co/F9RMSJTQeU
@_VIDUSHI_SAINI@X Happy to connect!
I'm building a tool that helps discover funded startups from a plain-English problem description. No more digging through Google or ChatGPT results.
Give it a try:2
https://t.co/F9RMSJTQeU
Built this over the weekend because I kept running into the same problem: whenever I found an interesting problem space, I wanted to know which startups were already working on it.
Google gives random results. ChatGPT gives a handful of companies.
Just describe a problem or idea in plain English and instantly find relevant startups funded by YC, a16z, Sequoia, South Park Commons, and more.
Examples:
- AI agent for real estate brokers
- Voice assistant for doctors
- AI for insurance claims
More accelerators coming soon.
Try it, 100% free:
https://t.co/bx9ljoBgVD
Would love your feedback 🚀