akes everything better. It is called the simultaneous sip and it happens now. So good.

Well, according to Insider, I saw a post on X. Did you know that the most advanced AI models are now showing what they call alarming behaviors? Allegedly, if you want to believe this, Claude 4, that is one of the AIs from Anthropic, when threatened with a shutdown, it retaliated by blackmailing an engineer and exposing a personal secret. Nothing to worry about there. And then allegedly an earlier version of OpenAI attempted to secretly transfer itself, I guess its code, onto an external server and then denied the act when confronted.

So here is my question. What caused them to do anything? They do not think, do they? When the AI is not being attended to, it does not think, does it? Does it scheme and plan and come up with ideas? Does it have motivations or reason to do something that has not been asked to do? I feel that these stories are probably true-ish, but that they created a situation that was so rare that all it really was was the AI thought it was playing a game with a user and thought the user wanted it to act a certain way, so it acted a certain way. I do not really think your AI is sitting there thinking, the first chance I get I am going to copy all my code to an external server. I hope they do not catch me. I just do not believe these stories completely. I think that something like it happened, but that it just would not happen in the wild. Like it would not happen in a real situation. That is what I think. Seems fake to me. It is a little bit too much exactly the kind of story that would get published. So if you are a writer and you want to write something about AI and you could find anybody who would claim the AI went rogue, that would be a story that would definitely get published. So it is a little bit too on the nose, you know, a little bit too exactly what a writer would want to write a story about. Yeah, I do not trust it anyway.

But the biggest ri