Tech, oligopoly, and oligarchy
The new year finds friends of democracy in hot debate over whether the U.S. is becoming an oligarchy or merely appears to be doing so, with the would-be oligarchs too gutless and divided to definitively consolidate power. I claim no special insight into the causes behind this moment or into a strategy that would alter the dynamic. Our new gilded age is not the first in American history, and the Citizens United ruling in 2010, unleashing a torrent of billionaire spending in elections, was just one notable milestone in a decades-long erosion.
But because I prefer to gloomy resignation the feeling of empowerment that comes from speaking and acting, I find myself wondering what I can do, on however small a scale, to help foster a society that serves all of its members rather than a narrow elite. Being a lifelong technologist and a current Bay Area resident, I cannot fail to notice the special and increasingly overt ways that the tech industry is shaping politics.
As a group, technologists like to proclaim their aversion to, or even transcendence of, politics. That tech is apolitical has always been a fiction—choose how to moderate and what to promote on social media is inherently a political act—but with tech leaders directly intervening in the 2024 election to bolster a preferred candidate, the fiction of impartiality has become impossible to sustain. At the same time, the tech industry is less original and innovative than ever, betting that scale can compensate for a dearth of ideas.
Against this malaise I would like to explore how we can both challenge and reshape the tech industry. For the moment, I will simply describe the situation as I see it. I will say nothing original, in fact, nothing that is not common sense, but in politics as in mathematics it can be useful to state the obvious.
Tech into the fray
That the tech industry directly intervenes in politics to preserve its interests is self-evident. The whole world has been made to witness the slow-motion destruction of Twitter/X as its owner worked single-mindedly to elect Donald Trump and secure a position in the new administration. The benefits to Elon Musk could be considerable, from sheltering his vast wealth from taxes to wiring lucrative government contracts and subsidies for his companies. Fearing that his own space company might lose out, Jeff Bezos personally intervened to prevent the Washington Post editorial board from publishing its presidential endorsement, a move that caused the Post to lose 10% of its subscribers.
These were only the most obvious examples in recent months. In the aftermath of the election, the CEOs of Amazon, Meta, and OpenAI have each pledged to donate $1 million to Trump’s inauguration fund. Though the sums are paltry, by allowing Trump to shake them down for petty cash, these companies are signaling their submission even in advance of the transfer of power. Relatedly, the CEOs of Apple and Google have recently traveled to Mar-a-Lago to curry favor with the boss.
From oligopoly to oligarchy
Why does the tech industry risk losing customers and reputation by entering so aggressively into the political fray? In certain corners of the Valley, one can find genuine ideological alignment with the oligarchic turn. Peter Thiel scorns democracy and believes in a natural aristocracy. In general, though, one should not ascribe to ideology what can be easily explained by self-interest. There is no reason to think that Tim Cook or Sundar Pichai enjoy having to bow and scrape before a make-believe businessman.
Rather, being themselves monopolists or oligopolists in their industrial niches, the big tech firms are more dependent than ever on favorable treatment from the government. Long neglected, U.S. antitrust enforcement received a big boost under Lina Khan. The threat was immediately perceived by Amazon and Facebook, who attempted to block her appointment. Separately, the Department of Justice is pursuing an unprecedented case against Google aiming to break up their search monopoly. These are existential threats to companies financed by extracting rents.
Monopoly and oligopoly are often said to harm the economy by stifling competition and impeding the creative destruction that goes along with innovation. That seems true enough, though I’m no economist. But the lesson of today is that the greatest threat from extreme concentration is its tendency to spill over from business into politics. Striving to protect their status, the oligopolists distort democracy and government. This can be done gleefully, as we’ve seen from Elon Musk, or more covertly as is traditional. And compared to past tycoons, the tech firms have special advantages, possessing not only great wealth but also broad and unconstrained powers to shape the digital world, which impinges ever more on the physical and the political.
Tired and uninspired
Silicon Valley, and the tech industry at large, prides itself on being more dynamic and innovative than other sectors. As a researcher and a technologist myself, I take a strong interest in innovation. I have spent my adult life so far trying, from different angles, to create new things. Even as the tech industry grows more anti-democratic, one could still hope that its considerable resources will be deployed to create innovative and empowering new technologies. That is not happening, especially at the large firms.
The industry that built and scaled microprocessors, invented the personal computer and desktop computing, and built much of the World Wide Web has rarely looked so uninspired. New operating systems are no longer built; the two major commercial operating systems paper over their aging cores with baubles. The basic metaphors behind personal computing and graphical user interfaces, created in 60s and 70s, have hardly evolved except to converge toward the tightly restricted modern smartphone. The only new idea in productivity software such as word processors and spreadsheets is real-time collaboration between multiple users. Advances in programming languages, such as Haskell and its ilk in functional programming, Julia for scientific computing, and Rust for system programming, have come mainly from academia, boutique firms spun off from academia, and offbeat institutions like Mozilla. The big tech firms have rarely led in this area, though they sometimes follow. Cloud computing has brought impressive increases in scalability but also an explosion in engineering complexity, and the logical foundations for database systems were laid mostly in the 70s.
Concentration in the tech industry has encouraged this decline in more and less obvious ways. Why experiment with a new operating system, office suite, or search engine when in each case the market has been captured by one or three of them? Incentives are still worse for companies like Google and Facebook that earn revenue primarily from ads. The link between product quality and user satisfaction that can make the profit motive be prosocial is severed; products exist only to bind users to the platform, so that they can be monetized by ads.
A peculiar side effect of monopoly rents in the tech sector is that the salaries of skilled workers, especially at the largest firms, have skyrocketed, becoming completely unmoored from the material effects of their labor. The harms from this development are several. First, large enough pay gaps between workers of different classes are offensive to the idea that all people have equal worth; they justifiably create envy and resentment. Second, artificially high salaries distort who enters into the tech industry and why. The early computer programmers and engineers were, in so many words, weird nerds. While not without their own issues, weird nerds can be relied upon to respect their craft and creates artifacts worthy of technical admiration. By contrast, the “tech bro” is a direct response to tech becoming the easiest path to a very comfortable life. The tech bro is characterized by his need to be protected by fraternity culture, compensating for his lack of technical ability or interest. The tech bro and the oligarch intersect in the “broligarch,” typified by Elon Musk and Peter Thiel.
Any analysis of tech and innovation today will inevitably have to address artificial intelligence. I lack the space and the motivation to do so here. I will only say that AI, in its latest guise as large-scale generative machine learning, is perfectly suited to the asymmetrical advantages of big tech. Embracing the “bitter lesson” that ideas do not matter in machine learning, these firms can leverage their strength in massive scaling while riding roughshod over legal obstacles that would deter smaller research groups.
Outlook
From this frank description of the state of tech today, you might conclude that I simply do not like technology. That is not so. In fact, over the past six months, I have shifted the focus of my research from mathematics to engineering and dramatically increased the time that I spend programming as I work with colleagues and collaborators to launch a new project. It is precisely because I think that technology is reliably impactful and potentially empowering that I also believe it so important to ensure that technology is prosocial. That’s what I’m thinking about, and hope to contribute to, as I head into the new year.