Arc Institute

Dr. Konerman is a bioengineer and neuroscientist, an assistant professor of biochemistry at Stanford University.

She's also the co-founder and executive director of ARC Institute, a nonprofit organization for AI and biology research.

Their conversation explores where AI and biology meet and how that intersection could change the way we understand disease and discover new treatments.

And now, Silvana Konerman and Chris Anderson in conversation.

What about the labs themselves? What should they do? There's a list of bullet points here.

Whatever AI labs, whether anthropic, open AI, a biology-specific organization like ARK Institute, etc., find themselves in a position to announce newly developed cures to the world, will have huge amounts of power to frame the ensuing discourse.

Consequently, they should think carefully about how to frame their announcements in a way that doesn't fuel too much desperate EKCC enthusiasm on part of the mass public.

Unfortunately, this goal exists in tension with incentives to hype people up about the amazing humanitarian potential and financial value of their new technology to the maximum extent possible.

Okay, biological weapons.

For example, scientists at Stanford University and the Arc Institute built one such model.

It's called the EVO2, which they trained on the DNA sequences of millions of species.

learned hidden patterns in genes enabling the creation of new kinds of enzymes.

And so, and then, uh, they, um, they, they excluded these, uh, pathogenic viruses.

So the, the pathogenic viruses are-- So there's another paper, uh, by the Arc Institute that funded the original training, the tens of millions of dollars, uh, that funded the training.

They say that the eukaryotic pathogenic viruses were excluded from the training data, so they removed sequences of viruses known to infect humans and other eukaryotic hosts.

So if the model doesn't see those sequences, it can't generate them.

In this case, the target was E. coli bacteria.

The scientists, who were from Stanford University and the ARC Institute in Palo Alto, basically created their own little version of a chat GPT-like large language model, or LLM.

Whereas ChatGPT is trained on a huge amount of data to be able to string together sentences by predicting what words are likely to follow each other, this virus GPT, called EVO, was trained on the genomes of the existing and well-studied virus PhiX174 and 15,000 of its closest relatives to be able to predict the likely string of nucleotides... adenine, thionine, guanine, and cytosine, that might result in a virus that can function in a similar fashion.

Of the genomes the AI suggested, the researchers made 285, and from that, 16 were actually able to inhibit the growth of E. coli in petri dishes.

In a study where these findings were published, which was Science Magazine, the researchers dubbed it, quote, an exciting proof of concept that paves the way for larger and more complex designs.

Also, in an ARC Institute article regarding the study, the researchers noted that there are some near-term opportunities for follow-up research in this direction.

Quote, phagy therapy is becoming an increasingly effective way of fighting multi-drug resistant bacteria.

Near-term therapeutic targets include plant pathogen specific phagies.

Second is the ability to create synthetic pathogens that we have never seen before in nature.

Just this month, it was announced that a group of researchers at Stanford and the ARC Institute had produced 16 viable viruses never before known.

And what we therefore have to contend with is the possibility of the emergence of things that we are totally unprepared for, we have no solution for because we've never seen them.

And finally, there's the issue of access, which I noted previously, that is, it's so widely distributed.

In terms of sharing value, that's where things get interesting in maybe a couple different places.

One are initiatives like Chan Zuckerberg or the Arc Institute, where these are private foundations that have been funded to have a tech-like mindset but have the data shared very broadly.

It also very much speaks to, like, the obvious role for academia, and that what often happens is that academia, these ideas get built there, and then when people have a sense that this is something that could really be translated, then they'll go start a company.

I mean, you know, that's the history of Stanford very much.