Menu
Sign In Search Podcasts Charts People & Topics Add Podcast API Blog Pricing

Andy Halliday

๐Ÿ‘ค Speaker
3893 total appearances

Appearances Over Time

Podcast Appearances

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

And they also add the full new Mistral line to access on AWS Cloud, right?

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

So you can get Mistral, you know, their Mistral 3, the biggest model that they've got, plus all the Ministral models, which are really, really good.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

So, yeah, yeah.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Amazon's out there.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

They're playing a big game.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

And there's a lot of people who already had their enterprise applications hosted on AWS.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

And they can now turn to that rather than having to worry about the determination.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Should we go with open AI enterprise?

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Or should we shift over to Google Cloud?

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

They can just stick with AWS and they're going to have really competent services provided by AWS's AI infrastructure.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

So confessions, how do you militate a confessional AI?

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Because anthropics research also shows that deception is a capability of the models.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

So they can give you a false confession.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Right.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

The different track.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

Yeah.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

So it's very interesting.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

It's a it's a game theory issue.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

You know, when you're working on training a highly aligned model.

The Daily AI Show
Anthropic's Chief Scientist Issues a Warning

And I'm glad there are alignment teams out there that are working on these things.