Jack Parsons was an important pioneer of American rocketry — a self-taught engineering genius who co-founded the influential Jet Propulsion Laboratory. He was also a bizarre sex cultist who was friends with Alister Crowley, the leading occultist, and L. Ron Hubbard, the founder of Scientology.
He would have made for a terrible babysitter. But you can’t deny that he knew a lot about rockets.
Sceptics of AI doomerism — some may call it pessimism — have been having a lot of fun pointing out the eccentricity of many of its leading minds. In these pages, for example, Andrew Orlowski has made sport of mocking the cultlike elements of the intellectual spheres that have been called the “rationalists” and the “effective altruists” — overlapping ideological-cultural tendencies that have tended to focus on exploring and promoting rationality, utilitarianism, and “longtermism” (thinking about our long-term future).
There can be a lot to mock about these spheres: the “we are all individuals” tendencies that make it likelier than you’d think for someone who is concerned about AI safety to be polyamorous and write like a cheerful automaton, the self-importance that results from their attraction to dramatic themes, and the intellectual absurdities that result from thinking narrow-mindedly from first principles. At its mildest, this can be merely entertaining. As its more dangerous, we end up with Samuel Bankman-Fried’s multi-billion dollar fraud, or the murderous rampage of the “Zizians” — neither of which was the direct result of rationalism or effective altruism but both of which reflected the arrogance and sheer antisocial weirdness that their more extreme tendencies can enable.
There are three points to be made. The first is that not all “rationalists” or “effective altruists” exhibit these characteristics. The second is that not all AI pessimists have even been involved with these tendencies. The third — and the more challenging — is that being weird doesn’t mean that you’re wrong.
This might sound obvious. (Wow, Ben, are you going to tell us next that there is something called the “ad hominem fallacy”?) But let me get a bit more grounded. A lot — though by no means all — of the opposition to recent pessimistic warnings about artificial intelligence have come from people sympathetic to the Trump administration, which has been adopting a decidedly optimistic and ambitious approach towards AI. Such people can look to sane, normal, grounded pro-AI figures like … Peter Thiel, the Antichrist-curious tech mogul who isn’t sure if it’s a good thing if humanity survives.
Some of the opposition also comes from the “e/acc”, or “effective accelerationism”, intellectual sphere. This tendency, which holds that unrestricted technological progress will lift human civilisation to new peaks, took inspiration from the early work of Nick Land — an English philosopher who took a downright favourable approach to machines replacing men.
Sometimes, you have to choose your preferred flavour of weirdo
To be clear, I’m not trying to say that you shouldn’t listen to Thiel, or e/acc advocates, or Nick Land because they are weird. I’m saying that when it comes to thinking about technology and the future, there is no non-weird option. Sometimes, you have to choose your preferred flavour of weirdo. Emerging technologies contain the potential — or, at least, the theoretical potential — for such novel disruption that both pessimists and optimists are liable to have eccentric tendencies to even entertain them. All kinds of outcomes can emerge from the same source. The hubris and heterodox inclinations that made Jack Parsons think “I want to join a pagan sex cult” also made him think “I want to let off massive rockets”.
Of course, there are a lot more eccentrics than there are eccentric geniuses. Most engineering-inclined odballs could not compete with Jack Parsons when it comes to engineering or to oddity. Of course, too, one can oppose a flavour of weirdness while welcoming another flavour of weirdness. (I love the weirdness of David Lynch’s films. I hate the weirdness of the Wachowskis’s films.) So, I’m not saying that pessimists about AI are right. I’ve been reading about AI safety, now and then, since Bostrom’s book came out and I still don’t feel qualified to have an opinion. What I’m saying is that when someone says “look at these weirdos”, there isn’t always a sane and balanced alternative lurking behind them. You don’t have to agree with weirdos all the time — indeed, you shouldn’t — but sometimes you have to say, “That weirdo has a point”. Arguably, technology and society have changed so much, so quickly, that “normal” doesn’t mean anything any more.