an N=1 exmple, but impressive if true www.youtube.com/watch?app=de...

Slightly diminish a band: SOSONOTBAD

Yeah that’s what I‘m thinking to do.

I have an af3 model and xtal structures of the individual proteins. In the model the two proteins are not significantly interacting, but I'm still wondering whether having two proteins fused could trip up MPNN 🤔

Is the science they do in industry different from the science you teach at your University? The level should be held high I agree, but expanding your horizon is a good thing. E.g. in engineering PhD programs, doing industry placements is quite usual. I think that's great and opens doors for people!

It'll continue as long as people have an incentive to get a PhD for an industrial career. Also many may think academia is their thing but reconsider once they're in that environment and that's ok. Unfortunately, academics tend to think there's only one right path that everyone needs to follow.

That's a bit short sighted of you. If you looked at a few job openings you find that many industry positions require a PhD. Before that's changing, people are forced to go that route. Profs should be open in supporting their students, whether you like the direction they choose after or not...

seems like the cybertruck came there on a truck? www.reddit.com/r/ForeignPla...

At least Elmo is considering himself a „longtermist“ and some of those definitely think along the lines of: ignore all current problems, divert all investments to AI which will then in turn solve all our problems…

Would it matter if there was a plan if the outcomes are the same?

Don’t fail to see the biggest pro: there is no actual freedom of speech on X. If your (benign) opinions are not in line with their ideology, you face harassment, persecution, and silencing by the platform itself (e.g. #democracy).

Something to help understand the "entirety" of literature would be amazing. However, it can put us in danger to produce even higher quantities of literature instead of focusing on quality. End-to-end paper generation comes to mind...

youtu.be/lkMWgP-b46Y?... another breakdown of AG1

The next one to watch… I hope the other funders have enough wits and influence.

It’s generally interesting that a Boltzmann machine trained on a 1000 sequence MSA could outperform this in terms of diversity… but I guess that’s just not as cool.

So many jealous people here that wish they were "smart" biofounder 😜