So why would anyone use OpenAIs API?
Especially now that you can use their models for much cheaper elsewhere. Look at this contrast. Both OpenAI models. One is a new reasoning model at over 2000t/s for pennies and the other one is their mini model for twice the price.

Today @giffmana.ai gave a talk at my University. Really nice to attend some events like this since I will not be able to go to Interspeech this year.

New OpenAI strategy: Let o3-mini con investors out of their money.

This was my first programming experience.

This is absolutely insane for this size

No country makes it so clear that rules are only for the poor and the government exists to serve the rich. Calling the Us a democracy is delusional at this point.

Predictions:
o1 release,
sora release,
some kind of update to the normal GPT,
A new feature in the app,
Maybe a open source model,
at least one blog post or paper,
Price reductions

I recently got my own Goban. It is so nice to play again. Btw if you haven't watched already I recommend the documentary about AlphaGo:
youtu.be/WXuK6gekU1Y?...

They also observe the same improvement with the length of reasoning.

Deepseek, the chinese AI company, is recreating o1 just months after the announcement. Soon we have this running locally.

I am pretty sure that o1-mini uses the base model of 4o-mini. This means it is very small (7-10B). Now imagine a open source alternative in 2025 that runs with thousands of tok/sec on Cerebras or Groq. This will reduce thinking time dramatically.

Time to get back into Go