Jonathan McRae

And I presume that is one of the real powers of having this idea of an AI scientist, that it can have such a broad understanding of the latest developments literally as they happen.

So the next obvious question is, do we actually want this to happen though?

Right.

So there are a number of commentators who would say, look, I think, you know,

the hype on AI has reached a peak and it just continues to grow.

The Gartner hype curve just doesn't seem to have the value that we normally expect with this.

So understand that there are other people in the area of AI who feel like a lot of this is overhyped.

But from your point of view, if we think about all of this, what about the concern of bias?

Because

While AI can be less bias in its examination of results, bias permeates all of science.

And I'm just wondering, how do you account for that when you talk about an AI that can read every single paper?

There's bias everywhere in language, in culture, and all of that permeates into science as well.

How do we get the AI to be better than humans at impartial science?

Yeah.

I mean, they are pretty convincing arguments.

It'll be really interesting to see where this goes.

The first peer-reviewed paper by an AI scientist has been published.

It's been really interesting speaking with you, and I'm sure we'll connect again over the next couple of years.

Researchers and scientists love to hear your thoughts on this work.

That's Jeff Klune.