Neil Freiman

My reading of the situation is that Amazon researchers tried to jailbreak this like many other tech companies did.

And Andy Jassy, the CEO of Amazon and Amazon, were just like raising general concerns about this.

I don't necessarily think that they were like, you need to cancel, you need to stop Anthropic in its tracks right now.

This is way too dangerous.

They were just like, hey, this is something that we found.

And then the White House took and ran with it.

And

And I think that because if you just look at what cybersecurity experts were saying about this particular jailbreak, they're saying like, you know what, it's actually not that huge of a deal.

This is not necessarily new to Fable or Mythos.

Other companies, AI tools can do this as well.

So they were not so concerned with the particular jailbreaking that Amazon was doing.

So I'm not sure Andy Jassy knew what he was getting into when he called up the White House and said, hey, I just want to let you know what we found.

Both sides have no idea what they're doing with their messaging because the U.S.

government at the same time is trying to restrict Fable and Mythos is allowing AI chips from NVIDIA, high-end AI chips, to go to China, which could ostensibly build their own model.

So they're trying to have it both ways.

And then you have Anthropic as well saying it would be good for the world to have the option for governments to slow or temporarily pause frontier AI development.

And then they complain when it actually happens to them.

So

We're in the wild west here of dangerous AI models, and people don't really know how both sides.

I mean, some people have ideas, but at least Anthropic and the government are in a state of chaos right now because both their policies and their messaging is incoherent.