How America's Energy Department is Building a National Platform for Doing Science with AI (acm.org) 33
America's Energy Department "wants to build a single national platform for doing science with AI," reports Communications of the ACM:
It is called the Genesis Mission, and the idea is to connect the country's 17 national laboratories, their supercomputers, scientific datasets, and a growing layer of AI models and agents into one system researchers can access. The DOE has taken to calling it 'a national operating system for science.' That means treating compute, data, and AI models the way the country treats power lines and highways, as shared national plumbing everyone else builds on top of.
If it works, Genesis will change how scientific work gets organized, checked, and scaled, with AI helping run the whole pipeline from hypothesis to simulation to experiment and back. The pitch is that this is better understood as infrastructure policy than as another research program. Genesis is now moving from announcement into execution. President Trump signed the executive order launching it in November 2025. This past February, the DOE published 26 science and technology challenges for the program, and in March it opened a $294-million call for research teams in fields like nuclear energy, quantum information science, semiconductors, and biotechnology.
The program is also beginning to reach beyond U.S. borders. In June 2026, Japan moved to become Genesis's first international partner. The two governments plan to invest a combined $1 billion over five years, with Japan contributing $500 million toward joint work in quantum technology, nuclear fusion, and biotechnology. The stated goal is staying ahead of China in the fields where AI is advancing fastest. The open question is whether a federated platform this big can actually work, or whether it ends up as one more expensive coordination exercise.
If it works, Genesis will change how scientific work gets organized, checked, and scaled, with AI helping run the whole pipeline from hypothesis to simulation to experiment and back. The pitch is that this is better understood as infrastructure policy than as another research program. Genesis is now moving from announcement into execution. President Trump signed the executive order launching it in November 2025. This past February, the DOE published 26 science and technology challenges for the program, and in March it opened a $294-million call for research teams in fields like nuclear energy, quantum information science, semiconductors, and biotechnology.
The program is also beginning to reach beyond U.S. borders. In June 2026, Japan moved to become Genesis's first international partner. The two governments plan to invest a combined $1 billion over five years, with Japan contributing $500 million toward joint work in quantum technology, nuclear fusion, and biotechnology. The stated goal is staying ahead of China in the fields where AI is advancing fastest. The open question is whether a federated platform this big can actually work, or whether it ends up as one more expensive coordination exercise.
the problem with a single AI platform (Score:5, Insightful)
"When everyone is thinking alike, no one is thinking."
https://quoteinvestigator.com/... [quoteinvestigator.com]
Re: (Score:2)
That's not a problem for them, it's the point. Thought-shaping and direct visibility into anyone trying to research anything. Not long until it becomes mandatory.
Re: (Score:2)
Well, we should certainly be able to distinguish between 'policy that has value for the community at large' and 'policy that supports a single stakeholder's objective (even when that's the stakeholder with the gold)'. So I don't disagree with you.
Re: (Score:3)
Well, we should certainly be able to distinguish between 'policy that has value for the community at large' and 'policy that supports a single stakeholder's objective (even when that's the stakeholder with the gold)'. So I don't disagree with you.
Or a single stakeholder who likes to cover things in gold.
Re: (Score:2)
Well you know you just can't get good gold paint these days.
Competition is good. (Score:2)
It's the Dead Poet Society march of consolidation.
Re: (Score:1)
Not only that. LLMs can, at best, give you low-hanging fruits. These are not very valuable and the ones that got overlooked will run out very quickly.
Explanation for Republicans. (Score:5, Interesting)
1) Science works.
2) Republicans hate government science, so they cut funding for it.
3) Democrats realize points 1 and 2 so they realize they have to get sneaky: Democrats take parts of the government the Republicans like and get THEM to fund science.
Which is why the Department of Energy is doing this rather than one of the several actual Science based agencies (for example: Federal Laboratory Consortium for Technology Transfer)
Note, we used to do this with the military all the time. The GOP refused to fund any computer technology, but they gave a crap ton of money to the Department for Defense. Which is why the internet was created by DARPA.
Man those republicans were easy to fool!
Oh no, we are not doing science, this is all about guns and bombs and shields. Yeah, this is not economic stuff, it is essential MILITARY defense.
Re: (Score:3)
I see it as the DoE was the most pliant bureau at this point in time. Harder to get the other ones to agree to this nonsense. Once the idea of submitting all your data directly to the government at end-of-day becomes normalized, then will come the mandates, and then restrictions on what you can study.
Wait, you thought this was about promoting science?
Re: (Score:2)
The people that proposed it thought of it as science.
What the government politicians plan on doing with it (and why they funded it) are an entirely different thing.
Re: (Score:2)
The whole "Genesis Mission" framing for this hollowing-out of our research institutions strikes me as nothing more than deflection, unless we also count it as a government handout to "regime-friendly" businesses.
"See, we aren't cancelling ALL science funding! We're building this! So shuddup about how we're quashing independent research and tearing out the ocean monitoring systems et al. and only giving federal moneys to people who already agree with us ideologically and will twist their results to fit our s
Re: (Score:2)
It had to be named after the Bible. Just to drive the point home that it's about science.
Re: Explanation for Republicans. (Score:2)
It's kind of true. The DoD proposals I've seen, even software, get labeled "weapons system" because Congress is more likely to fund it that way. In a very real sense, the DoD budget is the largest jobs program in America, and has been for decades.
This will be very effective (Score:4, Funny)
One of the problems America currently faces, is that we're still getting far too much science done, it's not costing us enough money, and the money it does cost is being wasted on paying the salaries of scientists instead of personally paying whoever contracts to kick back the most to political appointees.
I believe this will help solve all three problems.
Re:This will be very effective (Score:4, Insightful)
No, it won't. It is merely a way of getting government to own all the science in the country so that only "approved" science gets funded. You can read that as "Christian science".
Re: (Score:3)
If you read the article, this only affects the National Laboratories. The government already owns those. Private institutions are free to carry on as usual.
Re: (Score:1)
For now...
One bite at a time.
Re: (Score:2)
AI Has Brought Only Problems for Humanity (Score:2)
Re: (Score:1)
DOE is combining #1 and #2 to give us Nuclear Slop.
Government involvement is... (Score:1)
...cancer.
I knew this was coming, but falsely believed an optimistic fantasy that somehow AI labs could avoid it.
I have lost my optimism.
The game is over, the bad guys won, just like they always have for all of history.
Re: (Score:1)
yawn "gubmint bad"
a bit less punch to that statement when most of us (including 50% of /.) voted for the bad gubmint. its what you wanted so stop complaining about it.
Re: (Score:2)
Things have to get bad before they can get good again. If we don't figure out the national debt thing, it doesn't matter anyway though.
Is this how Big AI some sweet gets taxpayer money? (Score:2)
The open question is whether a federated platform this big can actually work, or whether it ends up as one more expensive coordination exercise.
It's big enough for big AI to fit in.
Genesis is now moving from announcement into execution. President Trump signed the executive order launching it in November 2025.
As long as he gets his cut.
Re: (Score:2)
Is this how Big AI gets some sweet taxpayer money? ...
Is this how Big AI gets some sweet taxpayer money? (Score:2)
Seems like I'm still sleeping.
Mission Genesis (Score:1)
once again business as usual (Score:2)
https://www.whitehouse.gov/pre... [whitehouse.gov]
How did Communist China get ahead anyway that the US needs to catch up or hold them off? We were all told capitalism
Good luck with chemistry (Score:2)
Re: (Score:3)
Both. You will need as many chemists as before, still AI will have great applications. And I guess many will leave US, but not because of AI, but because of politics.
Grift (Score:3)
Researchers within the DoE (and pretty much everywhere else) have been using machine learning in their work for decades. This kinda sounds more like a handout to various companies and reducing the number of tools researchers have access to.
Always the push towards Unity (Score:2)
This push is another aspect of organizations not being or wanting to grok anything unless they cram it into one black box. This appears to stem from observing something similar that happens in science. In physics, this is the push to get one gonzo-whopping theory that Explains All. In computation it gets expressed as Turing Machines are all we need. However, if you look closely, you find there are frequently many different ways of expressing whatever it is theory is trying to explain.
Political "leaders" and