Sam Fazeli

Just contrast the approaches, perhaps, or maybe...

Maybe it's too much for a podcast today, but I'd love to hear how you think everything fits together with regards to the current scene of protein design, AI-driven protein design.

That was brilliant, Mark.

Thank you very much.

Let me pick through this a little bit.

We have the cell, the protein target, and the data.

Right?

The cell disease cell.

Now, I've read Andrew White has made an estimate that you need 10 to the 38 flops to simulate a human cell.

Do you need, do you think, to simulate every single atom to be able to make an in silico cell useful?

Or can you do it, or do you not need to go to that level of perfection?

That's great.

Thank you.

I was hoping you would say that because I don't think we have enough energy to run something that requires 10 to the 38 flops of computation.

But so, Mark, on that specific point, a lot of diseases cause not, I mean, it's

The perturbation may be on a single cell level, but there's a network effect that comes in, in terms of an organ basis or paracrine, a variety of ways that the biology is more complicated than a single cell level.

Where do you think we are?

Actually, two questions here.

Where do you think we are in terms of those, what you just spoke about, in terms of having those models functionally useful already?

And where do you think we are in terms of trying to get the network