OpenAI’s new o1 model can use reason to solve math and science problems.
For the past few days the tech world has buzzed with rumors about Strawberry, a top-secret model from OpenAI with the ability to reason. The rumors got the name wrong, but they got the concept right: OpenAI released "o1" today, which they describe as "a new series of reasoning models for solving hard problems."
Here's OpenAI's description:
We trained these models to spend more time thinking through problems before they respond, much like a person would. Through training, they learn to refine their thinking process, try different strategies, and recognize their mistakes.
According to OpenAI CEO Sam Altman, o1 has already been rolled out to 100% of ChatGPT Plus subscribers.
It's funny the secret codename for this release was "strawberry" because it happens to be the first model that correctly answers the question "How many 'Rs' are there in the word 'strawberry'? ChatGPT 4o would usually answer "two"
That's lit
My first attempt to use it, it hallucinated a quote by Seth Godwin, but overall the output was of far superior quality. I'm excited to see where this is going.
"use reason" -> "use reasoning"
So wild.