Dr. Richard Moulange

But you could test how well the viruses are killing the bacteria.

And some of the best ones do better than the best existing bacteriophage we'd ever found before.

This is huge.

We can now design organisms, small ones, to do things better than we have ever seen in nature.

We can go beyond nature in this very narrow subdomain of biology.

And this heralds the promise of genome scale engineering, which is going to be, I think, a revolutionary capability within biology.

At the end of last year, there was a great paper from the AI red teamers at Microsoft, Whitman et al.

And it's recently published in Science.

And what they showed is that you can take

Current, off-the-shelf, open-weight biological tools.

So not necessarily these language models.

They might be sort of more specialized, a bit how AlphaFold is specialized for predicting protein structure, but it can't do every single biological task under the sun.

And they took ones that are instead good at generating protein sequences in particular and designing proteins.

What they did is they made lots of designs for ricin.

So ricin is, it's actually sort of two different proteins together in what's called a complex.

And it's a known chemical weapon.

Essentially, yes, ricin can be considered either as a chemical weapon or a biological weapon.

Yes.

You can derive it from living organisms.

I'm not going to discuss in depth how you can do that.