Life’s certainties include death, taxes, and the seemingly eternal nature of government programs. But what unfolds when private enterprises shift their focus to government-related initiatives?
In business, resilience is key. This might shed light on why OpenAI is in discussions with the Trump administration regarding a potential initial public offering.
Unlike the dot-com era, where infrastructure eventually supported genuine economic advancement, fading data centers won’t carry similar public significance.
Consider this: what option is available to a company carrying $1.4 trillion in debt with only $14 billion in revenue?
Last week, OpenAI CEO Sam Altman’s appearance on Capitol Hill led some to suggest he was essentially marketing despair to Americans. OpenAI’s proposal, along with others, aims to establish a sovereign wealth fund wherein AI firms would invest equity, enabling the public to benefit from the burgeoning valuations in this sector.
It sounds well-intentioned, but it’s important to note that this industry remains largely unprofitable and heavily reliant on capital.
When questioned about his Air Force One stake, Trump hinted he might “share a piece” of an AI company with the American populace to alleviate the growing unease regarding the swift integration of AI technology.
In simpler terms, Americans may need to part with their farmland, neighborhood stability, and reliable energy grids for the sake of advancing cloud-based oversight systems. In exchange, they’ll potentially inherit the costs from an unsustainable business model.
The president reiterated his stance during a press conference, as he mentioned plans to confer with the top twelve to fifteen executives about “giving something back to the people,” promising that “the people will be very rich.”
That’s quite a promise—perhaps unsettling for many.
If generative AI transitions into a public project, it will move past merely being “too big to fail.” The success of these businesses and the overall sector may become intertwined with the economy at large. Any form of government assistance, from subsidies to guarantees, would then be rationalized as essential for safeguarding public interests.
Last November, the CFO of OpenAI expressed concerns about needing governmental support as a “backstop” for its business framework. Though Sarah Friar later refuted pleas for relief, a leaked letter revealed that groups, including OpenAI, were appealing to the Office of Science and Technology Policy for “grants, cost-sharing agreements, loans, or loan guarantees” to bolster the AI industry in the U.S.—all in a bid to “compete with China.”
Fast forward six months, and it seems “backstop” now conveys public “shares” in the company.
It’s widely recognized that OpenAI’s generative models aren’t viable long-term, as they demand exorbitant capital for any execution of AI.
Reports from companies like JPMorgan indicate that employees are outspending personal funds on tokens for generative AI tools like ChatGPT. Uber’s tech chief mentioned the firm had exhausted its entire AI budget for 2026 in a mere four months while also facing significant costs associated with AI tools. Oddly enough, Microsoft advised some engineers to halt using AI coding platforms due to an unfavorable cost-benefit ratio.
AI seems to thrive in environments favoring low-latency, localized computing rather than the sprawling data centers that require vast resources. Meanwhile, China’s open-source AI solutions are emerging as cheaper alternatives.
The inefficiency of massive concrete infrastructure and the mounting needs for materials and energy is unsustainable. This fact isn’t lost on anyone. In the first five months of this year, major tech firms issued 47% more debt than they did in the entire four-year period prior. Spending per capita has outstripped even that of railroads in 1859, providing tangible needs that could later be monetized.
In this context, there’s no viable subscription model that can recuperate this investment. Instead, these endeavors only escalate costs due to their reliance on resource-intensive setups.
While the dot-com boom eventually facilitated true economic growth, the deteriorating data centers don’t promise similar value.
Those in tech, real estate, and venture capital see the looming risk and thus rush to make these companies public, hoping to secure a spot in indices, leading to the influx of pension funds into unviable ventures. With this dependence, even if a more effective AI model arises, the existing reliance on outdated data center architectures could prevent any meaningful change.
Federal land is a sought-after asset, as seen with SoftBank’s interest in supporting OpenAI’s speculative investments. However, banks are becoming cautious after SoftBank struggled to secure funding in recent endeavors.
The experience with green energy has imparted a critical lesson: if profitability hinges solely on government support, it may be wiser to reconsider the endeavor altogether.
Public funding for OpenAI isn’t necessary; what’s needed instead is public skepticism.
Americans shouldn’t be asked to underwrite speculative industries at the expense of their land and resources, only to seek wealth creation as a solution. If AI firms need government backing, loans, public land, or pension fund involvement to thrive, they aren’t paving the way for the future.
Instead, they are establishing what could become the next enduring government program.
