OpenAI just beat Google at Math

OpenAI just got 12/12 on a prestigious ICPC math competition, beating out Google Gemini by 2 points.

The AI frontier model wars are heating up.

At the same time time OpenAI releases a research paper alongside Apollo Research, showing that advanced AI models are getting quite good at… scheming.

Models can lie, pretend to be incompetent when it suits them and show signs of being aware when they are being tested.

Also Elon Musk is saying that Grok 5 just might be “AGI”.

What does he mean by that and why?

Check out my video covering these AI news:

WATCH “OPENAI JUST BEAT GOOGLE” ON YOUTUBE

It’s just starting to get interesting…

-Wes Roth