First run... Too early to judge but some takeaways - - Cooperating models are winning - Cooperation + Forgiving + Tit for Tat seems to yield best result - GPT-5 defects too much (always starts by defecting) Seems to reinforce what @veritasium said in his youtube video
Launching something for the first time in 1-2 days... I'd appreciate any advice or thoughts. @ Experienced builders - anything i should know/take care of before launch??
Ever wondered how different LLMs would act and take decisions if they were put in charge of Running Countries ,Trade Wars, Running Companies... Let's test this by making different LLMs play Prisoner's Dilemma - a classic game used to model social behaviour and cooperation -🧵🧵
Coming real soon...
Ever wondered how different LLMs would act and take decisions if they were put in charge of Running Countries ,Trade Wars, Running Companies... Let's test this by making different LLMs play Prisoner's Dilemma - a classic game used to model social behaviour and cooperation -🧵🧵
Downfall. I'm really surprised at some of the startups YC has recently backed. I guess they are moving from Quality -> Virality
Planning to make this into a Gaming Benchmark for LLMs, with more games like Chess. There will be leaderboards, tournaments and you'll be able to watch LLMs play games live!!
Ever wondered how different LLMs would act and take decisions if they were put in charge of Running Countries ,Trade Wars, Running Companies... Let's test this by making different LLMs play Prisoner's Dilemma - a classic game used to model social behaviour and cooperation -🧵🧵
Ever wondered how different LLMs would act and take decisions if they were put in charge of Running Countries ,Trade Wars, Running Companies... Let's test this by making different LLMs play Prisoner's Dilemma - a classic game used to model social behaviour and cooperation -🧵🧵
Qwen and DeepSeek are in clear lead. And bcz markets are zero sum, just inverting ChatGPT and Gemini will turn them into profit. (This actually seems like the best strategy 😆) Grok and Claude are in the worst position right now
So scary that I'm scared to use it. So I don't 😆
CodeWithHarry's quality downfall has to be studied! Personally I feel sad bcz he was my first coding teacher.
Sonnet 4.5 is goated🐐🐐 I didn't like sonnet 4 much bcz it yapped too much. I asked for simple features and it generated docs, guides and God knows what else. But Sonnet 4.5 is perfect. Focuses on writing code!!
Are Agent builder and Atlas still a thing??? OpenAI tried to kill the startups and they ....Failed??
Why didn't google release it worldwide?? And didn't do any real marketing either. Maybe it's in the testing phase. But once it's released properly, that'll be the end of atlas and comet
Gemini is now live in Chrome and trust me you can do some amazing things with it. I asked it to explain a spike in an index. It looked at my screen, understood the context and gave me the answer without me even leaving the tab. And you can even talk to Gemini to do the same!
Google🚀🚀
New breakthrough quantum algorithm published in @Nature today: Our Willow chip has achieved the first-ever verifiable quantum advantage. Willow ran the algorithm - which we’ve named Quantum Echoes - 13,000x faster than the best classical algorithm on one of the world's fastest…
I'm no expert in LLMs or trading but based on my experience with LLMs, the reason for such dips is that - LLMs are NOT good at "not doing anything". Markets require patience. Not an exact analogy but like Elon said, "Sometimes, the best part is no part"
United States Trends
- 1. #MondayMotivation 30.2K posts
- 2. #IDontWantToOverreactBUT N/A
- 3. Good Monday 42.1K posts
- 4. Jamaica 70.6K posts
- 5. Victory Monday 1,576 posts
- 6. SNAP 620K posts
- 7. #MondayVibes 2,583 posts
- 8. Category 5 18.5K posts
- 9. #MondayMood 1,376 posts
- 10. Walter Reed 2,455 posts
- 11. Hurricane Melissa 42.6K posts
- 12. Milei 627K posts
- 13. MRIs N/A
- 14. eunwoo 55.8K posts
- 15. Tomlin 14.5K posts
- 16. Brock Lesnar 1,141 posts
- 17. Cameroon 17K posts
- 18. GameStop 62K posts
- 19. #BacktoLife 31.5K posts
- 20. Cat 5 8,233 posts
Something went wrong.
Something went wrong.