The title of the executive order is on the short side for Trump: “Launching the Genesis Mission.”
It reads in part:
In this pivotal moment, the challenges we face require a historic national effort, comparable in urgency and ambition to the Manhattan Project that was instrumental to our victory in World War II and was a critical basis for the foundation of the Department of Energy (DOE) and its national laboratories.
According to Michael Kratsios, the science advisor to the president, that Manhattan Project comparison is just the beginning. The Genesis Mission is also, we’re being told, “the largest marshaling of federal scientific resources since the Apollo program.”
Then again, the Trump administration says stuff. The president said nuclear weapons tests were going to begin “immediately,” and that was almost a month ago.
But consulting AI.gov, Trump’s special fan page for showing off his love of AI, I find that the president has nine marquee AI executive orders, stretching back to his previous administration, and they have titles like “Promoting the Export of the American AI Technology Stack,” and “Preventing Woke AI in the Federal Government.”
None of them sound nearly as hauntingly mysterious as a “Genesis Mission.” What’s this AI-loving president up to now?
What the “Genesis Mission” is literally supposed to be:
We’re being promised a sort of AI and automation super-platform for the federal government. Based on my read of the program laid out in this order, the Secretary of Energy—fracking mogul Chris Wright—is supposed to unify all Department of Energy datasets with those of all federal agencies, and use those to create “scientific foundation models.” Presumably that means the government’s own LLMs, or other LXMs used for scientific research.
Then our federal government is going to use its new AI models to build programs that “automate research workflows, and accelerate scientific breakthroughs.” We’re getting set-it-and-forget-it federal science, in other words. The AI does the research, and a person can just come along and scoop up the breakthroughs like cream from a milk bucket.
According to Politico, Wright says there will be an “incredible increase in the pace of scientific discovery and innovation.” They’re looking at nuclear fusion, other energy sources, pharmaceuticals, protein folding—all the areas of science and research that pair well with AI hype.
And what does the plan specifically entail?
The executive order does, in all fairness, outline what the next year (and beyond) is supposed to look like for this program to some degree.
By the 60-day mark: A list.
America gets a document identifying 20 core science “challenges” the Genesis Mission can solve.
By the 90-day mark: An inventory.
America is gifted an inventory of computational resources the Genesis Mission can use to build its system.
By the 120-day mark: A plan.
By now, the Mission is supposed to have its data optimized and in place to train the models.
By the 240-day mark: Another inventory.
Wright is supposed to have figured out where robot-driven, automated science experiments can be done. Since it probably sounds like I’m joking, here’s what the order says exactly:
“Within 240 days of the date of this order, the Secretary shall review capabilities across the DOE national laboratories and other participating Federal research facilities for robotic laboratories and production facilities with the ability to engage in AI-directed experimentation and manufacturing, including automated and AI-augmented workflows and the related technical and operational standards needed.”
By the 270 day mark: A demo.
We get some sort of proof of concept for the Genesis Mission platform, focused on one of the 20 aforementioned challenges.
Within one year (and then every year from now on): An evaluation.
Were positive outcomes achieved? Did the Genesis Mission make scientific discoveries? How’s everything going? It’ll all be in the annual report.
And the Genesis Mission had better work, because the other side of this effort is a bunch of federal funding cuts for science. This administration has sought to cancel federal funding for (and subscriptions to) science journals. It has sought to cut $783 million in funding for health research—cuts that, it appears, really will go into effect. It has sought to cut off funding to no less than 100 climate change studies. It has reduced research spending at the National Oceanic and Atmospheric Administration by $100 million, and on, and on.
The cuts may have had many aims, at least one of which was to curb DEI (remember when people used to talk about DEI?). Another one, it seems, is to shift science into the realm of things you can just automate, well before sufficient AI systems exist to justify anyone’s confidence that such a thing is possible.
So buckle up for automated, low-cost scientific breakthroughs, everyone! They’ll be here soon, thanks to the Chris Wright and the Genesis Mission. Otherwise, some of those funding cuts might start to look a little silly in retrospect.
Read the full article here
