how far can an ai get by improving its own work? & how can we tell? write a starcraft bot in C → play → lose → rewrite ↻ games let us mea...

how far can an ai get by improving its own work? & how can we tell? write a starcraft bot in C → play → lose → rewrite ↻ games let us mea...

StarSkirmish is a mini RSI arena surprised me to see @OpenAI's Astra so far ahead https://twitter.com/kaimcpheeters/status/21110952274362246...

This is what we’re doing @goodstartlabs ! Some of our game publisher partners have already netted 7 figures from the major labs If you are...

Gaming is such an amazing space Requires technical, social, & design skills That expertise is truly distributed often driven by true passion...

More & more AI is given a goal & told to keep going Benchmarking AI on this auto-research, understanding tendencies & strategies, directi...

Sonnet 5.5 is a great model, fantastic tool use, fast. It's also highlights the tradeoffs different models make in this case, humor We saw...

Kai gets it. Games are the playground for intelligence. Since 2006 people have been using the Brood War API to write scripts to play & comp...

Awesome to see @EGafni & Diogo launch their first model Jev! Verification is the bottleneck for building great AI learning environments. ...

“make AI better at my job” is a data problem. someone has to turn what “good” means to you into data* a model can learn from. there’s an e...

Two years of AI progress. Still struggling to get the joke. On 14,000+ real @playbadcards hands, Astra tops our updated LOL-Alignment bench...

increasingly clear that 'data' companies (rl envs, human / synthetic data ...) need to train their own models, on their own data, to solve r...

Applying models to what you do, is becoming essential institutional infrastructure. Training them is increasingly becoming a part of that....

AI is moving from: learning by example, → learning from experience. another win for the bitter lesson 💊 https://twitter.com/shuchaobi/stat...

Few understand the power of wiki's as a primitive. literally designed for context management https://twitter.com/hammer_mt/status/2093342884...

highlights: • Training on diverse synthetic games improves science reasoning, code generation, & procedural reasoning while competition mat...

Highlights: • We want [AI] to represent our interests against a counterpart who has their own • qualities that make an assistant pleasant ...

While we’re at it can we get trends.claude & trends.chatgpt Would be fascinating https://twitter.com/jeff_weinstein/status/209191366796526...

Some highlights: • Since DeepMind’s foundation in 2010, the constrained yet rich worlds of games have played a critical role in understandi...

we get asked a lot, why AI & games? I love games Some test strategy, others humor, speed, trust. Simple rules lead to the craziest situati...