When it said “i think” 😳😭😳😭😳😭 #ai #chatgpt @chatgpt

I just tried OpenAI's new model, and I am at a loss for words. If you give this cipher, aka decoding codewords, to chat2pt4o, you're going to get a response that says, To make meaningful decoding, I would need more context about the transformations or letter shifting that might be involved. Could you provide additional information? Now if you give the same cipher to o1, you get this new button here that says show chain of thought. And when you expand it, it's wild. It's counting the number of letters in the cipher, it's adding up the total number of letters. This is my favorite. It goes, hmm, letters matching. This is great. It goes, wait a minute. I think maybe there is a, I think, I think, it's still going. It's it's still going. Imagine how much time it would take to pump through all of this math and trying all these different methods. And you can see how the model is literally thinking through and trying different methods to arrive at the right answer. Which it does, by the way. There are three R's in strawberry was the answer. Pretty incredible. And if you want to see a video on how reinforcement learning works, like what does that actually mean? How did we, how did we build this? Drop a note in the comments and I can make that video next.