Peter Thiel — tech billionaire, libertarian polemicist, Trump donor — lately gave a speech on the Oxford Union, one of many oldest and most prestigious pupil debating societies on the earth, to kick off its 2 hundredth yr. That’s hardly information — we’ve all heard Thiel’s spiel many occasions earlier than on campus conformity and the way solely tech can save us.
However my ears pricked up this time as he particularly criticized my discipline. I’m an existential danger researcher at Cambridge College, the place my colleagues and I examine the dangers from nuclear and organic weapons, local weather change, and rising know-how corresponding to artificial biology and synthetic intelligence. All of those applied sciences pose extremely excessive dangers — we suppose it’s believable that a number of of them may result in civilizational collapse or extinction, affecting everybody alive right this moment. As many within the efficient altruism neighborhood have argued, I believe tackling these dangers is a key precedence of our time.
Thiel appears to have had a passing curiosity in these matters a decade in the past, talking at some conferences and donating some cash. However to my data he has not engaged with the existential danger discount neighborhood for so long as I’ve been concerned. As a substitute, he appears more and more excited about seasteading and the alt-right.
So why was he criticizing the sector of existential danger discount? Thiel appears to counsel we in the neighborhood are Luddites, bearing some accountability for the stagnation in actual wages and technological progress for the reason that Seventies. He claims a number one reason for stagnation is that scientists successfully have develop into too terrified of their very own know-how. He informed the Oxford Union viewers that “the one reply as to why it’s stalled out on the a part of the schools is one thing like science and know-how are simply too harmful.”
Existential danger isn’t a hoax
I don’t wish to decrease the scenario. It truly is true that actual wages for a lot of employees within the UK and US have been stagnant for the reason that Seventies, particularly by way of the final grueling decade of austerity. And an excessive amount of technical effort and enterprise capital has been suboptimally invested into e-commerce (like PayPal), social media and internet advertising (like Fb), or surveillance (like Palantir). (What’s the connection? All three firms helped to make Thiel’s estimated $8 billion fortune.)
However as somebody who spends a good period of time encouraging technologists to think about their obligations for the know-how they create, let me say that an overabundance of concern isn’t usually what I encounter.
Highly effective applied sciences are sometimes “dual-use”: They can be utilized to assist or to hurt folks. Take nuclear physics. Nuclear weapons are in some ways the unique existential danger, the one which has loomed over the world since 1945. Nevertheless, nuclear energy can also be a dependable, zero-carbon supply of energy.
Advances in biotechnology enabled the quickest vaccine rollout in historical past, in addition to additional medical breakthroughs. However in addition they allow “gain-of-function” experiments the place scientists purposely attempt to make ailments extra virulent.
Language fashions like ChatGPT are great and superb however within the improper arms may allow the mass manufacturing of disinformation, as we warned about in “The Malicious Use of Synthetic Intelligence” in 2018. The issue with arguing for “extra pace” is that these dual-use applied sciences are already transferring quicker than we will sustain.
Thiel is solely improper if he thinks that “gradual it down” is the one response. The Founders Pledge Local weather Change Fund, a neighborhood of entrepreneurs who pledge to donate a portion of their exit earnings to charity, tries to hurry up innovation in low-carbon concrete and metal. Alvea is a biotech startup aiming to hurry up vaccine manufacturing. Anthropic is an AI security firm attempting to hurry up interpretability and alignment. We name this ”differential know-how improvement”: rushing up protected or defensive tech relative to dangerous or offensive tech.
But it surely’s additionally clear that the outdated Fb motto of “transfer quick and break issues” received’t work. Simply rushing up know-how received’t be sufficient to maintain us protected. We’d like smart home regulation: supporting the inexperienced transition, elevating security necessities in organic labs, and guaranteeing that high-risk AI programs undergo security checks. We’d like worldwide agreements, just like the Paris settlement and the nuclear arms management treaties that Donald Trump — whose presidential marketing campaign Thiel donated to — ripped up.
A libertarian method to X-risk doesn’t work
However Thiel doesn’t appear to need this. He’s an Ayn Rand libertarian. On ideological grounds, he doesn’t imagine that authorities motion can assist. He thinks regulation makes issues worse. When requested by an Oxford Union viewers member about how he “would repair” the UK’s Nationwide Well being Service (NHS), he stated we have to recover from our “Stockholm syndrome” as a rustic and privatize it already. Everyone seems to be entitled to their very own opinions, even billionaires. But it surely’s a fringe ideology. And it’s one that might do an enormous quantity of hurt.
Thiel seemingly adopts this fend-for-yourself mentality in his planning too. He has lengthy had a bolt-hole that he may escape to in case of societal collapse. In 2015, he purchased a 193-acre plot of land (greater than the Disneyland theme park in California) on the South Island of Aotearoa New Zealand for a reported $13.5 million. In Could 2022, he was denied planning permission to construct a luxurious lodge on the plot with house for twenty-four visitors, a theater lounge, a spa, and a “meditation pod.”
A mate of mine was touring across the South Island a couple of years in the past and had a pint at a pub. One of many locals identified a spot on a hill: “See there? That’s Peter Thiel’s home. If something goes dangerous, that’s the place we’ll go for meals and water.”
It’s a fantasy to suppose that existential danger will be decreased — or survived — simply with particular person motion, the invisible hand of the market, and a “go quicker” signal. It additionally wants collective motion, knowledge and endurance, and smart and proportionate regulation.
That’s the method that the existential danger and efficient altruism communities are taking. However that, sadly, is the method that Thiel seems to disagree with.
Haydn Belfield has been tutorial undertaking supervisor on the College of Cambridge’s Centre for the Research of Existential Threat (CSER) for the previous six years. He’s additionally an affiliate fellow on the Leverhulme Centre for the Way forward for Intelligence.
