Marc Andreessen's Mindless Techno-optimism

Plus: Iran’s “imminent” attack, dangerous AI agents, Silicon Valley’s “Oppenheimer”, perverse US-China spat, and more!

Tell me if I’m romanticizing the past: Wasn’t there a time when the titans of American capitalism were less annoying than they are now?

I mean, sure, they considered themselves kings, entitled to all the power and privilege that befits a king. But did they consider themselves philosopher kings? Did they issue proclamations about metaphysics and social theory and the broad currents of world history?

This question occurred to me some months ago, when I was listening to Marc Andreessen, billionaire co-founder of the venture capital firm Andreessen Horowitz, share his thoughts on the mind-body problem with podcaster Lex Fridman. Though it’s true that Andreessen has a mind and a body, I still didn’t understand why we were turning to him for enlightenment on this particular issue. 

And why do we listen to Elon Musk wax ontological, telling us that the chances we’re in “base reality” (as opposed to a simulation) are one in a billion? Or listen to Peter Thiel expound on the cultural forces that keep us mired in technological stagnation—when, as any fool can see, we’re not mired in technological stagnation?  

But let’s get back to Andreessen, because his bigthink has taken the form of online treatises that seem to have actual influence. In particular, they are motivational texts for people in the “accelerationist” camp of the AI debate. There’s Andreessen’s “Why AI Will Save the World,” from last summer, and his “Techno-Optimist Manifesto,” which came a few months later. Both are works of such sweeping assertion, and such sparse elaboration, that their persuasiveness depends on the reader granting Andreessen’s stature as a major thinker, a master of fields ranging from evolutionary biology to economics to world history.

Another thing these essays have in common is their upshot: Government regulators should keep their hands off of technology and let it evolve in accordance with the laws of nature, which, if given free flow, will apparently carry us all to some super wonderful place.

If you dissent from this techno-optimism, says Andreessen (again and again), you are part of a long line of doubters who, in the past, have always been wrong. He writes in “Why AI Will Save the World”: 

The fear of job loss due variously to mechanization, automation, computerization, or AI has been a recurring panic for hundreds of years… We’ve been through two such technology-driven unemployment panic cycles in our recent past – the outsourcing panic of the 2000’s, and the automation panic of the 2010’s. Notwithstanding many talking heads, pundits, and even tech industry executives pounding the table throughout both decades that mass unemployment was near, by late 2019–right before the onset of COVID–the world had more jobs at higher wages than ever in history.

Now, let’s think about this. The concern underlying the “outsourcing panic of the 2000’s” was that jobs would migrate from high-wage countries like the US to low-wage countries like China, thus reducing employment and/or wages in the high-wage countries. This is a concern about effects at the national level—about good economic effects in China having as their flip side bad economic effects in the US. And Andreessen thinks he can evaluate the validity of this concern by giving us statistics about the global economy? America’s working class couldn’t have suffered because in the end “the world had more jobs at higher wages”? 

That doesn’t make any sense!

The fact is that, as pretty much any economist will tell you, outsourcing did hurt many American workers, especially in the manufacturing sector. And, by the way, so did automation. In fact, lots of economists say that automation has played an even bigger role than outsourcing in taking jobs from working class Americans. And for various reasons—such as the fact that replacing workers with machines makes more economic sense in high-wage than low-wage countries—automation, like outsourcing, is something whose effects aren’t well illuminated by Andreessen’s global statistics.

I’m actually pretty pro-technology. And I’m a big booster of international trade—which often has subtle and underappreciated benefits (like lower prices for American consumers) and, I think, tends to reduce the chances of war. And I certainly recognize that both globalization and technological advance tend to increase aggregate prosperity in the long run. Still, short-term term effects at the local and national level can be important. For example: 

The effects of outsourcing and automation on America’s working class helped get Donald Trump elected president. (Rhetorically, Trump ignored the effect of automation and focused instead on the effects of outsourcing and immigration, since demonizing foreigners is his stock in trade.) And the populist revolt that swept Trump into office culminated, as you may recall, in the storming of the US Capitol—an attempt to violently overturn a presidential election and subvert the US constitution.

To those of us living in the real world, things like this matter. But from Andreessen’s Olympian vantage point, I gather, they are mere ripples in the larger currents of history. And if only the rest of us understood these currents, we would understand that everything will work out well in the long run so long as the government doesn’t raise taxes on billionaires or inflict burdensome rules on the kinds of companies Andreessen invests in. We would understand that (as one plank of his techno-optimist manifesto asserts) “everything good is downstream of growth.” (Not just some good things—every good thing.)

Also, by the way, “Energy is life.” And this fact leads only a few sentences later to this proposal: “We should raise everyone to the energy consumption level we have, then increase our energy 1,000x, then raise everyone else’s energy 1,000x as well.” That may sound ambitious, but actually it’s just the first step. “We should place intelligence and energy in a positive feedback loop, and drive them both to infinity.” (Why stop there? I think we should go, along with Buzz Lightyear, “to infinity and beyond.”)

All of this abstract theorizing, disconnected from the real world, coming from an unelected and unaccountable billionaire who is insulated from the negative consequences of his ideas, would be at least somewhat easier to take if it weren’t for this passage in his manifesto: “Our enemy is the ivory tower, the know-it-all credentialed expert worldview, indulging in abstract theories, luxury beliefs, social engineering, disconnected from the real world, delusional, unelected, and unaccountable–playing God with everyone else’s lives, with total insulation from the consequences.”

This faux populism is one of the more annoying things about Andreessen. And, like Trump, he is happy to infuse his populism with the kind of conspiratorial air that isn’t exactly something America needs more of right now. His techno-optimist manifesto begins with these words: “We are being lied to. We are told that technology takes our jobs, reduces our wages…”

In the real world, and in real time, technology does take jobs and reduce wages. Yes, technology can also, in the long run, give displaced workers better, and better-paying jobs. It has often done that in the past, and it may do it again. But if the nearer-term disruptions get big enough, that long run could take a very long time to materialize. Maybe America can survive a populist revolt that’s bigger than the last one, but I’m not eager to find out.

Some of us believe that artificial intelligence is about to bring a burst of change so dramatic as to have unprecedentedly high potential for social destabilization. And we think its disruptive effect on employment, dramatic though that will be, isn’t the half of it. We’re not against the full unfolding of AI, and we see the great benefits that may bring, but we’re against it happening too fast, with too little reflection and guidance. And we think the only way to keep it from happening too fast involves government intervention—and, ultimately, coordinated intervention by governments around the world.

And that, I guess, is what I would call the techno-realist manifesto, or at least a start on one.

Maybe it’s a wrongheaded manifesto, and these concerns are off base. But if any techno-optimists want to convince me of that, they’ll have to do it from somewhere other than Mount Olympus.  —RW

PS I don’t mean to suggest that Andreessen’s entire argument for AI remaining essentially unregulated is the generic argument that technological progress is good in the long run. His pro-AI manifesto has a slightly more granular argument than that, and I will address it in a future issue of the newsletter.

This week, psychologist Paul Bloom came on the Nonzero podcast and talked about things ranging from his total-eclipse experience to the legacy of the recently deceased psychologist Daniel Kahneman to the question of whether “western enlightenment” can save the world without help from “eastern enlightenment.” During the Overtime session, Paul and Bob discussed whether AI could also play a role in saving the world. If you don’t have a paid NZN subscription, and thus don’t have access to the 47-minute Overtime segment, you can still watch this excerpt from it at NZN’s new Nonzero Clips YouTube channel. Feel free to click the like button—or to go even further and subscribe to the channel.

As the world awaits Iran’s retaliation for the Israeli air strike that killed three Iranian generals last week—retaliation that the New York Times says is “widely expected to be imminent”—there’s a chance that developments in Gaza could avert it. 

Trita Parsi of the Quincy Institute points to signs that Tehran is looking for a way to reconcile two hard-to-reconcile goals: (1) avoid retaliation so strong that it could start a wider war; (2) not lose face. According to an anonymous Arab diplomat cited in Iranian media, Iran told the US that it will attack Israel directly unless President Biden secures a ceasefire in Gaza. The White House has denied the claim, but if it’s true, writes Parsi, it suggests that Iran “wants to avoid a direct confrontation with Israel, but it cannot avoid it unless it secures a big win in the region.”  

Parsi also speculates that, "Tehran may suspect that a ceasefire is already in the offing, allowing it to use that as a pretext to both take credit for ending the war and avoid getting into a shootout with Israel." In any event, now that this report of an Iranian ultimatum is out there, a ceasefire would make it politically easier for Iran to forego retaliation or at least opt for minimal retaliation. “The long list of strategic reasons why Biden must secure a ceasefire has just grown longer,” concludes Parsi.

Also in Responsible Statecraft, Paul Pillar, former US national intelligence officer for the Near East and South Asia, suggests that if Iran indeed manages to avoid severe retaliation, Benjamin Netanyahu may be disappointed. Pillar writes that the Israeli assassination of Iranian generals looks like “an effort to escalate Israel’s way out” of its Gaza-related difficulties (difficulties such as: global opinion turning against Israel, American support wavering, and the aim of “eliminating Hamas” seeming out of reach—not to mention Netanyahu’s domestic unpopularity and his legal challenges). 

By provoking Iran to retaliate, Pillar says, Israel could 1) portray itself as acting in self-defense, diverting the world’s attention away from its increasingly problematic assault on Gaza, and 2) drag the US into conflict with Iran, further involving Israel’s superpower ally in a regional crisis that could otherwise become a threat to Israel’s security.

This week, the Biden White House sent a strong message to China: Stop helping us fight climate change! 

That’s not exactly how Biden officials put it. During Treasury Secretary Janet Yellen’s trip to China, she just suggested that China should stop flooding the global market with cheap clean-energy technology—solar panels, electric vehicles, lithium-ion batteries. 

The Biden administration complains that these exports are “artificially” cheap, meaning that prices are held down by government subsidy. But isn’t that one way you’re supposed to fight climate change? Isn’t the whole problem that market forces alone won’t save the day, because the cheapest forms of energy are often the most carbon-intensive? So we need governments to either make that energy more expensive by taxing it or make the cleaner alternatives less expensive with subsidies? 

The Biden administration does have a plausible concern: