dekhn

Simple chemistry couldn't be considered intelligent- it's basically just falling down a potential energy gradient. It gets more interesting when you implement chemical logic (like transistors) and graphs of molecules that convert between each other, like autocatalytic networks.

See also: https://cyberneticzoo.com/cyberneticanimals/1969-the-logo-tu... In my gifted & talented class in the 80's we were introduced to computers and I remember seeing the picture of "Irving" and thinking "Wow, I should learn LOGO to program robots when I grow up" (unfortunately, recursion scared me so I switched to BASIC and later assembly).

What do you mean "prozac is fluoride"? Its chemical structure contains a trifluoro group, but it's not fluoride... and you don't end up with extra fluoride in your system.

As a side note: I take prozac and it's far better than paxil (for me and many others). Has been transformative in my life.

Going off of paxil was absolutely horrible, and at the time I did it, the psych community didn't believe that you experienced side effects during taper or discontinuing.

I have never in my life felt worse than when I was getting off paxil. Beyond the head zaps, it pervaded my intellectual with depression and anxiety.

Only after a few "head zaps" and other papers got published did the psych folks even listen to me about it, and then they were just sympathetic and tried to find me something else.

Thalidomide is still prescribed.

AlphaFold didn't solve a decades old protein folding problem. They substantially improved protein structure prediction but with important limits.

THe primary bottleneck for a drug is getting it through clinical trials. No matter how much you speed up lead detection, development, and in vivo studies, the bottleneck will still be clinical trials. Now, if we could make lead detection better at generating molecules that get through clinical trials, that's great- but it would take decades to know for sure. If we could make development less likely to produce toxic molecules that trigger the immune system, sure that's great- but again, we're talking decades before we know if it worked.

Attempting to improve clinical trials- well, those are already fairly well designed and run, but it's likely you could make improvements there too that increased efficacy. But again, you're still just dealing with the FDA, filing tons of paperwork, and tons of people looking at the paperwork. Unless you fix those systematic issues, we're not going to accelerate drug approvals.

Yes, we explicitly avoided any kind of high concurrency/parallelism. That was an intentional decision, although note we did have a global torrent network across all the workers, whose topology (node interactions) formed an extremely complex graph.

I used to work in the MPI world on supercomputer jobs and it was miserable. I also worked on other high concurrency systems at Google (helped run the Ads DB, did training and engineering of Sibyl). I continue to insist these systems are incredibly simple compared to even the most basic human cells interactin g.

The system I built was called exacycle. It ran as a borg job in every cluster we had access to, with priority 0 or 1, with N tasks (where N=# of machines in the cluster). Borg ran the "worker" task as long as there were resources free on the machine (often, several workers ran on the same machine, and no workers ran on other machines). There was a global schedulers that picked mini-tasks off a global queue and fed them to the cluster-level scheduler (batches based on instant resource availability, and subject to some data locality optimization) which then fed them to the machine workers.

At peak, we had 1-3 million xeon cores running our tasks (protein folding, protein design, drug docking), 24/7. This was before GPUs were really a thing. Each mini-task was basically a rosetta or gromacs or autodock binary running on local inputs and writing local outputs for ~1 hour using ~1GB RAM and no network. Input and output was via colossus, or distributed through a global torrent network (literally all 1-3 million cores formed a single data transfer group, with data passed between clusters at low probability).

I built most of the code of the system (which depended on Google infra) including the scheduler and the machine worker, and also ran the system as on-call "SRE" for 3 years. During that time, we nearly almost didn't take down the Google global network, and had small latency impacts on many services. Managing to run the system for so long without antagonizing ads or search or youtube was a major challenge.

I would say the complexity of the system I built (considering everything down to the OS) was roughly the level of a few human cells in a plate, in terms of interacting components and the complexity of their interactions.

Here are some of the publications that came from running exacycle (we provided strong evidence that using more compute improves outcomes in molecular simulations)

https://pubmed.ncbi.nlm.nih.gov/24345941/ https://pubmed.ncbi.nlm.nih.gov/24265211/

Many lab experiments involve waiting for a period of time, and automating the steps before and after would be both expensive and failure prone.

I interpreted the question more broadly, replacing "diesel" with "fuel".

AlphaFold's goals were twofold (heh): demonstrate that protein structure prediction (for a subset of proteins) was solvable, and to accelerate drug discovery. As most people in the pharma field already know, PSP was never the gating step in getting drugs approved.

I've done SWE work at the highest level (I ran a job that ran a worker on almost every idle machine at Google, and the workers formed a global torrent for data distribution). If you took all of Google's data centers in the world, along with their networks, it would be roughly the complexity of a few cells in the human body. Fortunately, with my training in biology, working on such simple data centers was pretty easy.

I just have to point out: thalidomide is still prescribed as a drug for cancer and is very effective.

Depends on context and the location of the road surface and the nature of the explosion.

https://www.sfgate.com/bayarea/article/Tanker-fire-destroys-...

(Also, the rebuilding was epic, the only time I've seen the state and a contractor move quickly: https://longreads.com/2024/07/15/rebuilding-the-maze/ )

Robotech (the americanized version of Macross and some other anime) was broadcast on regular TV (local stations) in 1985. I watched the entire series (I was 12 at the time) with my jaw on the floor. Already The Last Unicorn had shown us what Japanese animation was capable of, and Robotech brought things up a notch (not in terms of animation quality, but in terms of story). Note that Robotech and the original Macross and other series are a bit different. We were watching akira by 1989 or 1990.

I distinctly recall that Robotech had a cross-dressing character well ahead of its time (for american audiences).

I clicked thinking this was going to be a paean to Robotech.

I like the DMD (https://en.wikipedia.org/wiki/Digital_micromirror_device) in DLP projectors: millions of tiny mirrors each of which is electrically controlled to move to a specific angle quickly on torsion hinges.

basically: some aluminum extrusion for structure, some linear actuators for adjusting the position, and a nice camera with a nice objective. Plus he took apart the printer to get the sample in frame and controllable.

Check out some early drawings by Santiago Ramon j Cajal; https://daily.jstor.org/cajal-imagining-the-brain/ (late 1800s)

Look at the background of those researchers: Demis was cogsci phd (and he is a general intelligence that can work on a wide range of scientific problems) and John Jumper has a masters and phd in chemistry. Their advisors and coauthors included some of the best-known ML people who worked on protein structure prediction previously. Personally I think they should have got the biology one, but they weren't unqualified scientifically speaking.

I wonder if the best statement we can make for future AI discoveries is that, ideally, they will allow us to move up the Kardashev scale faster.

The new faster multiplication algorithms are both galactic, are they not?

It does not seem surprising to me at all that AI can find new galactic algorithms that are not of any direct use (at best they can guide us to find better solutions that are non-galactic).

What would be surprising and interesting is if AI can find a new floating point multiply-add that is 2X faster per operation than the current fastest chip implementations. That would immediately have billion+ dollar implications. The teams who already work on this have excellent approaches and invest enormous amounts into tiny improvements already; there are probably not much low-hanging fruit there (proponents of using AI for chip design improvements have impressive claims, but unclear that any of those really worked out in reality).

biological material of humans? you'll have to expand your argument here. Are you reconstituting the material on the other end? If you have that technology, do you need a robot to tell a human the story of earth? Why not just upload it directly?

This is where I start to get curious.

Why haven't these theories had more impact on materials science? We use DFT and other greatly simplified approaches (with tremendous results) but still I suspect we are massively underperforming in terms of producing materials with useful properties.

This is a cop out argument.

I've worked in ML since 1994. I've been very impressed with the progress of models lately. I read "The Coming Technological Singularity" when it was first written and dedicated my career to seeing these sorts of things coming true.

That doesn't mean that their exponential improvement will change faster than light travel. Therefore, to deliver a human being to the surface of another planet (let's say Proxima Centauri at ~4 ly away) with enough resources to create a stable colony means we'll travel at less than the speed of light. Even with a solar assist, the best we can do right now is less than 1% the speed of light in an unmanned probe. People are working on alternatives, but those are not the sort of technology you just pop in a space ship and send for 400 years to a neighboring star. How do you support a population for 400 years?

As an alternative, let's just make humans like tardigrades, capable of shrivelling up and surviving high-radiation, low-moisture environments. Then we can take 10K years to get to a remote location. As a bonus, the resulting humans can survive in a much wider range of conditions so we have many more target systems to consider. Oh, but that's human engineering- are you sure that we'll just be able to genetically modify people (technologically and morally) at that level any time soon? A single person died in an early gene therapy trial and it shut down the field for 10+ years.

Basically, what you're arguing is that exponential growth in AI (not a given) will unlock huge material sciences and biological advances. While possible, it's unlikely to combine in a way that suggests "we should send people to other solar systems".

Huh? I'm not sure that makes any sense.

Nobody is solving colonization of planets in this lifetime unless we find a way to accelerate mass faster than the speed of light (and decelerate it at the other end) or extend the lifespan/radiation resistance or can preserve bodies in suspended animation for long periods. Even then, that just removed one of the many technical hurdles. We don't know which planets to target- are you going to send people to solar systems when you don't know if they are habitable? Modify people to be able to inhabit otherwise hostile worlds? How do you deal with boredom on a 1,000+ year trip?

Even with aging, our efforts are better spent increasing quality of life across humanity than helping a few rich people live to 150.

Is there any obvious path from these results to any sort of physics, chemistry, or biology breakthrough? From what I can tell: no (but I lack the physics and chemistry expertise to say anything about the QCD finding).

Turning the question arouhnd: what would be the most impactful thing for OpenAI or Anthropic to do to show large breakthroughs?

I've had static binaries on linux machines that worked across distributions for decades.

I've seen several "mine cathedrals" now: https://en.wikipedia.org/wiki/Salt_Cathedral_of_Zipaquir%C3%... and a few more listed here: https://www.historyhit.com/beautiful-subterranean-salt-mines...

Isn't he the guy who posted about his LEGO-sorting computer vision project? That post alone taught me things that increased my career opportunities significantly.