Whatever they may be, the real reasons for SoftBank's $500 B investment aren't the stated ones
By now, we’ve all witnessed the great pomp and fanfare with which the the world’s largest financiers, the upper echelon of tech companies, and the US government announced that they are getting together to build the largest supercomputer ever built. My first reaction was, “do they really have to call it Stargate?” I wanted to know how Elon Musk felt about this, so I Grok’ed it:

Seemed about right. OK, so there’s this massive military program that will require nearly all of the solar system’s resources so we can be part of an intergalactic civilization then? Turns out, no—per Larry Ellison, it’s a cure for cancer … using personalized mRNA vaccines. This announcement begs for at least its face value to be questioned. Because I enjoy doing that sort of thing, here are my thoughts. (My thoughts are continually in motion, so subscribe to stay with The Focus Group!)
The Claims are Completely Disingenuous
In my experience, highly funded projects with moonshot claims are never for their stated purposes. The stated purpose is not the intended purpose. But the intended purpose won’t sell to a general audience, so investors fabricate stories to socialize their activities. To prove it to yourself, feel free to try out my prompt on your LLM of choice:
What are some examples where the stated purpose for a very expensive project turned out not to be true?
Whether for high speed rail in California, a bridge to nowhere in Alaska, a football stadium with a crushing bond, or an effort to save the world’s children from hunger (what Milan Kundera calls The Grand March), just ask yourself and your AI Agent:
Who benefits from such a project, regardless of the success of its originally stated purpose?
Usually, the actual reasons for such elaborate projects turn out to be mundane: somebody needed to get paid (greed), needed to curry favor (corruption), or wanted to control the human race (dominion). Let’s investigate the publicly stated claims of Stargate.
The Investment Produces Jobs
This is completely bogus. If your goal is to create jobs for Americans, building data centers is among the least economical and worst way to go about it. It doesn’t take DeepSeek-R1 to tell you this: every LLM knows. But I love how ChatGPT o1 phrases its answer, so I’m providing its response:
Data center construction and operation often yield fewer permanent jobs per $1 billion of investment relative to many other industries because data centers tend to be capital-intensive rather than labor-intensive. You might see on the order of a few hundred full-time equivalent (FTE) jobs (direct + indirect) per $1 billion of data center investment once operational, plus a larger but temporary number of construction jobs (e.g., 1,000–2,000 over 1–2 years). Other market opportunities, such as certain types of manufacturing, infrastructure, or service-sector investments, typically generate higher job multipliers. For example, $1 billion spent on traditional infrastructure or labor-heavy sectors (like education or healthcare) might support 2,000–10,000 direct and indirect jobs—often significantly more than a similarly sized data-center project.
Let’s also remember that the ten data centers in Texas listed as part of this project are already under construction, so those construction jobs have already been created. Likely, the remaing data centers will be built by the same crews.
The Energy Grid will Improve
No; no it won’t. President Trump said he’ll do whatever he can to make it so data centers can produce their own power, so very little of that energy will make it to an upgraded grid. And after draining energy resources from American soil, it’s more likely that these data centers will end up in catastrophic disasters, like uncontrollable fire at the battery facility in Moss Landing, CA earlier this month.
We’ll Cure Cancer
Of course we all want cancer cured, but will this thing cure all human ailments? And just because Oracle acquired Cerner recently doesn’t mean a new data center will suddenly cure cancer. Also, did Larry seriously just say mRNA vaccine?
The related claim about how rural and tribal doctors can gain access to diagnosis, skills and knowledge from the world’s top doctors is equally problematic—if only because they’ll need high speed internet, computing devices, and medical facilities, as well.
As an aside: I live in Utah near Huntsman Cancer Institute—how much of this investment is going to that fine institution? That would add a whole lot of jobs around here and would actually cure cancer.
Keeping AI in America
Just because we build data centers in America doesn’t mean we’re keeping the tech, the entrepreneurs, the money, or anything else in America. Exhibit A is the recent announcement of DeepSeek-R1, which rivals and even exceeds OpenAI o1’s performance. DeepSeek even accomplished this feat at much lower cost while open sourcing the entire model. We live in an age where the world’s largest companies aren’t just multi-national, they are themselves among the largest nations on the planet. It’s anachronistic to talk about oligarchies within nation-states. Spot the world leaders in this picture:

Following the Money
Ground Zero
Some of the funding will certainly go to a company SoftBank invested in called LayerZero Labs, the primary developer behind Stargate Protocol. The Stargate Protocol enables fluid movement of crypto tokens across blockchains to make lots of money. Stargate was officially launched in March 2022 and shockingly attracted a Total Value Locked (TVL) of over $1 billion in its first week, sparking widespread attention in crypto circles. Clearly, lots of people in crypto believe the Stargate protocol needs to become the infrastructure underpinning a broad swath of DeFi and Web3 applications across blockchains.
The Surveillance Society
Many observers immediately pointed to Larry Ellison’s previous statements about his desire to use AI as a method of keeping people on their best behavior. For me, the term “best behavior” is highly nuanced, but Larry believes it’s behavior that those in power think you should be doing. It requires a LOT of cameras, recording devices, databases (such as Oracle’s) … and data centers. So, yeah there’s a whiff of dystopian self-interest when he says things like this.
Backroom Deals
We may never know all the private deals being made in smoke-filled rooms. We can be certain they have nothing to do with the stated purposes of the Stargate announcement, namely curing cancer, creating American jobs, keeping tech in the US, and improving the US energy grid.
Final Thoughts
I’m happy for all the people who will benefit from this deal. I experience authentic joy in seeing natural competitors finding a way to sing Kumbaya around a fire together—even if it’s an exclusive fire. But what we don’t need is that same chorus representing that they’re doing all of this for our betterment, even at the great sacrifice of ourselves. I would just remind everyone to remain skeptical and vigilant.