Cora 🐬 retweeted
NVIDIA did it again..
they trained a GPT that generates human movement instead of text.. next-token prediction, but the tokens are body motions.
it's called GPC and it hit a 99.98% success rate reproducing a massive corpus of motion clips.. cartwheels, vaults, flips, all of it.
→ 99.98% success rate reproducing motion clips
→ learns cartwheels, flips, vaulting from raw data
→ adapts to new tasks with <1% extra parameters
studios spend 6-12 months and millions of dollars building custom controllers for one character in one game.
GPC does it in one forward pass.
Pixar, Ubisoft, and every mocap studio on the planet just felt a chill down their spine.
the same next-token trick that gave us ChatGPT is now doing parkour inside a physics engine.
Cora 🐬 retweeted
Claude Fable 5 has produced a hand-checkable counterexample to the Jacobian conjecture, an open problem dating to 1939.
The conjecture says that a polynomial map with a constant non-zero Jacobian determinant must have a polynomial inverse.
An 87-year-old problem was casually solved by Fable 5 on a sunday evening. Insane times to live in.
hello there the jacobian conjecture is false thanx to my close friend akhil for asking about it and my other close friend fable for working during the world cup final
((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z): \C^3\to \C^3, has jacobian determinant -2, and sends (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to (-1/4, 0, 0)
Cora 🐬 retweeted
Published this article 3 days ago.
Today it's:
#2 on Google
Showing up in Google's AI Overview
Didn't expect it to move this quickly.
Makes me pretty optimistic about where LLaMaRush is headed.
Cora 🐬 retweeted
90 days of building apps
here's what I learned 🤝🏻
Apps
- design a viral feature/screen first
- have viral video ideas before you even start coding
- mascots are still a good anchor for apps
- don't overload your app with features, it's okay to keep it simple
- study what's working, if you have an app idea, check if there are similar apps and if they're doing well
Marketing
- ASA gives $100 for free, use them to test
- localizing apps in ASC helps a lot with visibility (and you can market locally)
- it all depends on your app, if your viral feature is designed well, it'll multiply quickly
- rn AI images + self recorded videos are a free way to get good videos
- having a tt account in your target niche + country is gold
- have multiple channels, test different things
- do slideshows + videos (on different channels) and double down on what's working later
- treat SM like a lottery.
- your video got only 80 views? That's still 80 people.
People
- having people to motivate you changes everything
- getting feedback from people is essential, they see something you missed completely
- ask family & friends or people on X
- check feedback, don't implement every single feedback immediately, evaluate for yourself & your app
Cora 🐬 retweeted
Definitely seems like mass producing iOS apps and hitting a few winners is the best route atm.
B2B SaaS can be good if you know where to find users. This seems to be getting harder as more and more "software solutions" are being released by anyone and everyone that touches AI tools.
B2C is really hit or miss because you have to really hit a solid painpoint or FOMO type users to really succeed there.
Cora 🐬 retweeted
Them: how do I grow my app?
Me: run ads
Them: I don’t have money
Me: then post 20 reels and TikTok vids a day
Them: that sounds like a lot of work
Me: it is
Them: idk if I can do that
Me: awesome sounds like I don’t have to worry about competitors then.
Welcome to business. It’s hard :)