Brandon (Host)

speaker

390 appearances 3 recordings 1 series first heard Jun 2026 last heard 16 Jul

Brandon (Host)’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 3 in all, peaking in Jul 2026 with 2.

Appearances

newest first · ▶ plays the moment

Latent Space: The AI Engineer Podcast · 🔬 The Coolest Diffusion Research Isn't in LLMs — Evan Feinberg & Sergey Edunov, Genesis Molecular AI · 1 Jul 2026 podcast

And now you are switching and you are saying, we are actually going to have automated systems which make decisions and then advance.

First of all, you did allude to you've kind of hit maybe the threshold to, you know, the like last November threshold where it just magically became useful.

So maybe that's part of it.

But are there other reasons for going in this direction?

And like, you know, why now?

Oh, I was going to say, how many can they learn and how many do they use?

Because I've seen that they like to use lots of them, but maybe they're not necessarily good with... But it's really hard, right?

So I wouldn't interrupt you when you say agent, then some people like using the term agent to just mean something automated, but other people

use agents to specifically mean you have an LLM, which is responsible for orchestrating decisions.

I mean, outside of biotech, it's universally the latter.

But sometimes in the space of bio, people are still using agents to refer to any sort of decision-based system, which is automated out of humans.

And maybe some of them are even more like traditional rules-based systems or something.

You're saying that you can take a crystal structure, which I guess in this case is framed as a graph or like a three-dimensional image that mole star is interpreting, I mean, in some way, and then it actually can make decisions upon what is binding to what and make hypotheses.

Like you've gotten to that point where you have seen it making effective decisions.

Yeah.

So if you historically have treated humans like you were making tools for humans, so now you have agents.

How do you reimagine the interaction of humans and agents and the decision-making process to go?

Is this you hope to essentially automate humans out?

You want to scale up and do more hypotheses?

Or, yeah, and then also the kind of the skeptic would ask, how do you also tell, make sure that your agents are not making weird decisions where they tweak the wrong hyperparameter and now you give you the human a bunch of...

Showing 281–300 of 390 · page 15 of 20