99% People don't know: Thinking power gap between different AI models is MASSIVE: Anthropic Claude Opus 4: Up to 32,000 tokens Google DeepMind Gemini 2.5 Pro: Up to 32,768 tokens OpenAI O3 pro: ~100,000 tokens, theoretically, never specified in docs.

Cannot believe: Everyone waited for DeepSeek R2 for 6 months in 1st half 2025 Now keep waiting for OpenAI gpt 5 for rest of 2025

New algo is weird

from what i've seen its mostly devs telling other devs to use claude code then it spreads like wild fire 😂

Wow 👀

What a wildly interesting piece by Prue prue0.com/2025/06/11/rar…

Who won the AI release war in last 24 hrs?

Didn’t know that!

ppl like opus 4.1 this much???

Self reflection: bored a bit by writing only AI-related tech/product tweets. Maybe just the recent months been quite boring. There are new releases but they are marginal improvements instead of disruptive. We are also working on pivoting from nanobrowser to more vertical

now we have a tie

heard same

this is the one i missed!

take a guess 👀

lol why can't he

On going. 12 hrs left!

Best time to swim is now