Morozov’s paean to “ecological reason” is a breath of fresh air, demonstrating how the Cold War perverted not only AI development but our capacity to imagine alternatives—technological forms untethered to markets and the military. He lays special emphasis on the way AI today captures the ethos of neoliberalism; I’d like to expand on the way financialization helps it do so. The funds being funneled into the generative AI boom reflect a particular array of interests and externalities, and behind it all looms a long-standing asset bubble underwriting the expansion of our global computational infrastructure.
In 2023 venture capitalists invested billions in startups chasing AI but were outspent two to one by just three tech firms: Microsoft, Alphabet, and Amazon. If this year is any indicator, that trend will hold steady: venture capitalists have raised tens of billions while Microsoft, Alphabet, Amazon, and Meta notched $106 billion in capital expenditures aimed at expanding their own AI infrastructure and capabilities. Early this year, OpenAI founder Sam Altman was pitching a $7 trillion plan to investors to build global infrastructure for semiconductor production, hyperscale data centers, and energy sources for both. Meanwhile, an increasingly lucrative alliance has emerged between Big Tech and fossil fuel companies; the former have been signing cloud computing and generative AI deals that maximize productivity (and emissions), while the latter build new coal and natural gas power plants to satisfy the exploding energy demands of new AI data centers. As these various actors—Big Tech firms, venture capitalists, egomaniacal founders, and the fossil fuel sector—egg each other on, ratcheting up investment to the trillions, asset managers like BlackRock and Blackstone are angling themselves to profit handsomely off the prospective deal pipeline.
Morozov is right that many of those who warn about a technology bubble tend to think it is going to pop any day soon. There’s no reason to think so: we have been waiting for the other shoe to drop for well over a decade to no avail. For a brief moment in 2022, a series of deflations and demolitions in the tech sector—along with the end of quantitative easing—suggested the era of ballooning valuations was over, but the correction has proven illusory.
During an asset bubble, it is only a matter of time until moonshot projects and zombie firms and business models sustained by misallocated capital finally see their demise. Can we afford to wait around? Given the success of firms like Uber, Lyft, and Airbnb in leveraging misallocated capital into political power that then reshapes markets and urban governance into forms that will sustain them once the capital gluts retreat, I would say no. Things are even more dire considering the glut of capital is being used to build out infrastructure, goods, and services that are hastening the collapse of our ecological niche. Any plan that seeks to promote ecological reason will indeed need a radical political project. It will have to tackle Silicon Valley and its financiers, Wall Street, fossil fuel companies, and now the defense industry—all at once. How exactly we break the back of this unholy alliance is unclear, but we can tease out the shape of some things we need.
Venture capitalists, their funds, and their portfolio firms enjoy a great deal of their power from the inflation of startup valuations. High valuations let firms use investor funds as an anticompetitive weapon to assail markets and states; they also let venture capitalists orchestrate lucrative exits. Draining this vast pool will require a tax regime that disadvantages this type of capital ownership with aggressive taxation or by shifting the ability to value assets (such as an equity stake) to, say, the IRS. Biden’s billionaire tax, which follows these contours, spurred a hysterical response and outpouring of support for Trump from the sector that many found surprising. As both Morozov and Ben Tarnoff observed, however, VCs have never been liberal stalwarts—they’ve been primarily concerned with preserving their ability to transform speculative gains into real wealth that then confers political and economic power. Weathering ridiculous responses to a policy proposal is one thing, but cobbling together support that survives a Silicon Valley eager to flex its lobbying chops will be another.
There will also need to be some sort of public alternative to technological innovation driven by venture capital. Cornell law professor Saule T. Omarova, Biden’s aborted nominee for Comptroller of the Currency, offers a blueprint for options that might be palatable enough for capitalists. Chief among them is a National Investment Authority (NIA) that provides public equity to public projects. Omarova’s vision is to build out infrastructure for financing green energy projects that provide high wages and ventures that insulate the country from supply-side shocks or fight against inequality one way or another. The NIA would also function as a public asset manager that can negotiate or coordinate the provision of emergency credit, take equity positions in failing or bankrupt firms, and restructure them in alignment with a public development strategy.
This public-spirited agenda could go further still. The federal government’s relaxation of the so-called “prudent man rule” in 1979—allowing pensions to invest more heavily in VC funds—spurred the industry’s growth from $100–200 million during the 1970s to $4 billion by the end of the 1980s. Reversing its ability to access pension funds should be another priority. With some resistance, the Securities and Exchange Commission has proposed rules aimed at tightening regulation of private funds by VCs, hedge funds, and private equity, but what else can we do to take advantage of this underbelly (and convince others to join us in doing so)?
A public investment option combined with an asset manager could also be used to experiment with market and non-market interventions. Aspirationally, we could seize assets like computational infrastructure and either spin them down or operate them publicly to drive private firms out of business. Meanwhile, intangible assets (datasets, algorithms, and so on) could be used to furnish an alternative research agenda that promotes ecological reason. The goal should not be to replicate the utilities model—markets and states dominated by public versions of Google and BlackRock—but to clear the land of obstacles that prevent us from pursuing genuine experiments.
Is there a way to advance such a project despite the capitalists and efficiency shills who will surely be aghast at the long-term prospect of diminished power? There just might be, starting with Morozov’s rousing call to arms—a reminder that in technology criticism we have fallen for the trap of documenting each sin of a seemingly impervious Leviathan, when the point is to change the world!
