So the AI models are getting bigger and more complex with each iteration of in-house builds. Is AI now writing the parameters? Can it beat itself at Tic Tac Toe and be smart enough to not launch...
So the AI models are getting bigger and more complex with each iteration of in-house builds.
Is AI now writing the parameters?
Can it beat itself at Tic Tac Toe and be smart enough to not launch nukes?
We've had "AI" strategies to tune various models parameters for a loooooong time. I still don't think it's "smart" or demonstrating any kind of thinking.
We've had "AI" strategies to tune various models parameters for a loooooong time. I still don't think it's "smart" or demonstrating any kind of thinking.
I thought MS had a special partnership with OpenAI? Seems strange to me that they would be competing directly now.
MS doesn't own OpenAI. I think it would be foolish of them, from a long-term perspective, to not pursue their own in-house models.
So the AI models are getting bigger and more complex with each iteration of in-house builds.
Is AI now writing the parameters?
Can it beat itself at Tic Tac Toe and be smart enough to not launch nukes?
We've had "AI" strategies to tune various models parameters for a loooooong time. I still don't think it's "smart" or demonstrating any kind of thinking.
Last time I checked, ChatGPT couldn't even remember the board correctly for tic tac toe 🤣