Trump Media & Technology, the parent company of Truth Social, on Thursday announced a merger with TAE Technologies, Inc., a leader in fusion technology that’s backed by investors including Google, Chevron, and Goldman Sachs. The deal, valued at $6 billion, creates the first publicly traded fusion company.
It managed to surprise the market, a rarity during a Trump presidency whose political brand has been built on shock value. The media company’s unlikely entry into fusion is ostensibly a bid to capitalize on the artificial intelligence boom and its skyrocketing power demands, which have pushed both investors and startups into a race for new sources of firm power. Fusion — which relies on a still unproven technology and has yet to be commercialized — could be transformative as a carbon-free, nearly limitless source of baseload power.
Devin Nunes, a Trump loyalist who left Congress to be CEO of Trump Media in 2022, said in a statement that fusion power will be “the most dramatic energy breakthrough” since nuclear energy in the 1950s and ensure the country’s AI supremacy, a major Trump priority. Since January, the administration has issued a flurry of executive orders aimed at beating China at the AI race, such as directing federal agencies to speed up permits for data center construction and preempting state laws regulating AI.
Despite the advances in fusion power in recent years, the technology is still unproven. A joke among skeptics is that fusion “is thirty years away, and will always be thirty years away”; many scientists predict the mid-2030s is the most realistic timeline for fusion power to enter the grid.
The goal of fusion is to harness the nuclear reactions that powers the sun to create clean, nearly limitless energy in power plants on Earth. Scientists have been working on this for decades, but historically the process uses more energy than it produces.
A breakthrough arrived in 2022, when the National Ignition Facility at Lawrence Livermore National Laboratory managed, for the first time, to break even on energy for fusion ignition. That program cost billions of dollars, and took decades to reach so-called “ignition.”
No company has taken things the necessary step further and managed to create a reaction that produces more energy than it consumes — a prerequisite to commercialized fusion. Many are trying, including TAE, Helion Energy (which has a power purchase agreement with Microsoft), and Commonwealth Fusion Systems.
The latter is considered the richest fusion startup in the world, having raised over $2 billion to date. CFS in September signed a PPA with the Italian energy giant Eni for a planned fusion plant in Chesterfield County, Virginia. The agreement is worth more than $1 billion, though the specific financial terms weren’t being made public.
The details of Trump Media-TAE’s project were sparse as well. The combined company said it plans to site and start construction of the world’s first utility-scale fusion power plant (50 megawatts) in 2026 — assuming it can secure approvals — and eventually scale up to 500 MW plants.
I'm a scientist watching a lot of very talented people who have dedicated their lives to solving those problems, now having their careers destroyed and deemed obsolete because they can't magically turn the work they do into something that can be done by AI. Because AI doesn't do any of that.
How does rolling out data centers make any of that a reality? We're nowhere near the point of AI actually accomplishing any of that stuff. Yet it's being jammed down everyone's throat anyway, consequences be damned.
Automation can save time, but it doesn't actually solve any of those problems. Many of these giant AI companies don't even have simple automation down. They're just relying on mass sweat shop labor.
Several Nobel prize winners disagree. You seem to forget AlphaFold, the rise of automated theorem provers in mathematics, the incredible advances in climate prediction and more all in the last 4 years
You seem to forget the people who are pushing for the removal of regulations in order to roll out data centers as fast as possible have submitted reports (likely AI generated) that are riddled with debunked climate science as a justification.
I'm not saying AI doesn't have an important role to play. I'm saying we're not at the point where there is any reason to believe the ends will justify the means being used to get there.
There seems to be a very real possibility we are speeding up the destruction of the planet to roll out a bunch of very expensive and energy sucking holes in the ground that will be empty buildings by the next decade.
There is great possibility in the future of AI. However, we are also currently doing a lot of harm by making funding cuts to basic science and clinical research in order to reallocate this money to "AI" because of what we hope to accomplish in the future.
NIH funding cuts have affected over 74,000 people enrolled in experiments, a new report says
NIH shut out hundreds of young scientists from funding to start their own labs
These people are so blinded by the glowing promise of what they hope to accomplish, they will justify all kinds of horrible shit to reach that point. They believe so strongly that as long as they approach the problem with a utilitarian mindset, then they are essentially saviors of humanity while ignoring all the unnecessary harm and destruction they cause in the process to help the "greatest number of people." Mainly because they want to profit and they want the legacy of getting credit for "saving humanity."
For example, even SMNRs are not without their own unique risks, especially in places that face severe weather and are prone to flooding. Yet in my own state (Louisiana) people want to cut safety regulations and roll them out right away. It doesn't matter if we are extremely prone to flooding, and we can't even go more than a few months without boil water advisories, the promise of what could (hopefully) be is too important to do anything other than completely throw caution to the wind.
The reality of science is that it can't always be exciting and innovative, there is no single savior or group of powerful saviors, and great science is definitely never the result of demanding it happen as fast as possible. Science is often very tedious and boring work, and there are some ways AI can definitely help with that, especially with big data. However, most of the tedious work still has to be done in order to learn what actually works and what doesn't. Not to mention it's always done with the understanding that real world variability can never fully be accounted for in a lab (or an AI model). That's why you have to proceed with reason and caution before charging full speed ahead.
This is also why many scientists choose academia and basic science over a more lucrative career in industry. The incentive of basic science is to learn and discover, not to increase the bottom dollar of the people who dumped their money into your work like they were placing a bet, and now they expect a return on that money ASAP, even if it turns out what they invested in doesn't actually work right away, or if it does more harm than good.