Legitimate cries for regulation of A.I. have been coopted by bad-faith actors that want to influence eventual regulation in order to create favorable conditions for themselves, says none other than Marc Andreessen, the influential venture capitalist and tech industry thought leader who cofounded Andreessen Horowitz, also known as a16z.
In an interview with a16z general partner Vijay Pande on the Bio Eats World podcast—sponsored by a16z—the billionaire VC laid out his theory regarding the A.I. doomsayers. Andreessen compares A.I. regulation to calls to ban alcohol during Prohibition, expanding on a parallel he previously drew in a 7,000-word manifesto on A.I. he published in June.
During Prohibition, calls to ban alcohol came from two groups of people, according to Andreessen: what he calls the Baptists, devout Christians who truly believed alcohol was the source of America’s moral decline; and the Bootleggers, organized crime that stood to make themselves incredibly wealthy selling black market liquor if alcohol became illegal.
It’s a rudimentary historical analysis, but potentially a rather apt analogy, which attempts to draw a distinguishing line between earnest lawmakers and watchdogs that want to ensure a new technology is properly regulated by governments, and large companies that want to exert undue influence on the process to create a set of circumstances that benefit themselves under the guise of protecting the public interest. (For what it’s worth, Mob historians are generally unified in judging that Prohibition was a turning point and the birth of modern organized crime in the United States.)
“Big companies basically want a government-protected cartel,” Andreessen says on the podcast. “They want basically a regulatory barrier, where to be in the A.I. business, you need to have fleets of lawyers and policy-affairs people. You need to be paying government officials huge amounts of money, and hiring huge numbers of people out of government and going through all these compliance checklists.”
A spokesperson from a16z declined to provide examples of who Andreessen was referring to when he said “big companies.”
To make his point, Andreessen cites the example of the Dodd-Frank Act, passed in the wake of the 2008 financial crisis. The legislation was ostensibly supposed to ensure that the then-struggling banks weren’t too big to fail, but Andreessen argued that it ended up being written in large part with the help of lawyers and lobbyists, resulting in a virtually unchanged landscape of mega-powerful banks.
Experts will likely quibble with Andreessen about the particulars of his analysis, but it’s undeniable that the same group of titans dominates Wall Street today as a decade ago, and many experts share the broad strokes of his argument about the ineffectual legislation. On the other hand, the recent regional banking crisis found weaknesses far away from the “too big to fail” landscape in the banking system, and Fed Chair Jerome Powell has talked about “the natural human tendency to fight the last war.”
According to Andreessen, the goal of onerous compliance policies is to stifle competition, not just from particularly successful and innovative startups, but also from the public at large.
“The goal of that is to basically make sure that there’s no sort of competition,” Andreessen says. “And in this case, it’s actually even worse than that, because the lobbying effort is not only to protect the big companies against startups, it’s also to try to protect them against open-source.”
Regarding potential monopolies in the tech world, Andreessen was a board member and early investor in Facebook, which was eventually the subject of lawsuits in 2021 for allegedly agreeing to an especially favorable advertising deal with Google. Critics argue the deal was meant as a handshake agreement to preserve both companies’ dominance in the digital ad market. Andreessen remains on the Meta board, despite some reported disagreements with CEO Mark Zuckerberg, as in 2016 over India’s net neutrality policy and later, around Facebook’s settlement with the FTC over the Cambridge Analytica scandal.
On the podcast, Andreessen was also particularly vehement in his opposition to bans on open-source coding for A.I., calling it an attack on free speech.
“You’re literally telling people, they can’t write code and put it on the internet, which I think is just totally beyond the pale,” Andreessen says. “There are policymakers in DC that are very worked up on this issue right now and they’re trying to figure out how to ban open-source A.I.”
In May, a series of tech executives, including OpenAI CEO Sam Altman, testified before Congress about A.I. Many of those executives went on to call for regulations, welcoming the idea of government oversight. Andreesen acknowledges this as an odd situation in which the usual dynamics of regulators imposing guardrails that companies try to skirt was flipped on its head.
“Usually regulators trying to regulate something get fought,” Andreessen says. “In this case, it’s like the wolf is being invited into the henhouse. They’re all fired up.”
In May, Senator Michael Bennett (D-Colo.) introduced a bill to create a federal agency to regulate A.I, signaling that lawmakers are following suit as Andreessen and others expected. Meanwhile, last month the European Parliament passed a draft of a law known as the A.I. Act that would be one of the first to regulate artificial intelligence.
