At an 80th birthday party at the weekend, I met an academic who was evasive about his field. When he finally disclosed “computer science”, I asked him why he hadn’t wanted to say, and he replied: “Because I cannot have one more conversation about AI.” I couldn’t ask him why not, because of stupid manners; that would have been one more conversation about AI. But I don’t want to have another conversation about AI either.
Nobody’s opinion, whether utopian or dystopian, seems to keep up with the thing itself, so everything has the laggy, outdated feeling of a BBC Radio 4 afternoon play about AI. There was one last week, and I listened to it patchily, thinking: “If AI had written this, it would have made a more sophisticated evaluation of the threat posed by itself, and been less hammy, unless the instruction had specifically been, ‘Write some dialogue in the style of a pretend-family on a party political broadcast from the 90s.’”
There are cheerleading bystanders, the people who trust that technological advance is generally productive and good. Rather than engage with any of the crunchy reality of this vast terrain, they’ll tell you instead that every discovery in history was scary at first, and yet was a slab on the crooked, miraculous path to enlightenment. Invariably, there’ll be a bit in the middle where I’ve stopped listening and started thinking about space being pointless, then wham, by the end AI will definitely cure cancer. This is often quite plausible, by the way. There’s a resigned dehumanisation, as if discovery were a thing we consumed, as if we could outsource intellectual adventure and merely reap its rewards, putting no dent in our delight or meaning. But then what do I know? I’ve never discovered anything. And you can hardly say to someone whose cancer is uncured: “But I wanted to have a crack at fixing it myself.”
AI’s detractors cover a huge spectrum. You know who’d be great at summarising them? Gemini – from the metaphysical (“This is the end of the human creativity”) to the deeply credible (“This is the end of the service industry”), from the broad and scary (“This is the end of the entry-level job, and therefore, of the prospects for the next and every subsequent generation”), to the old-fashioned leg-work case (“Here are 17 graduates whose applications have been rejected within two seconds of receipt, because they were evaluated by AI, which has already taken the job of the HR person and has most probably also taken the job they’re applying for”). There’s a question mark over the billionaire overlords and whether or not their interests are pro-social, to which the answer is so self-evidently “no” that it feels lame to mention it, but does it follow that we should stop mentioning it?
There’s a case for the environmental harms of AI so glaring that it’s too depressing to look at directly, and there are intersections of military and politically repressive AI potentiality that feel like standing on the edge of a completely novel abyss. And all of it urgently needs consideration, but every conversation is stalked by the knowledge that in the time it took you to discuss it, AI has already become better at doing the thing you had haltingly realised it shouldn’t be doing in the first place.
The very act of chitter-chatter feels like surrender. This is what will sink us: our lust for deliberative, easily diverted, often inconclusive yakety-yak, in which there are so many cross-currents. Often we’re not even trying to find an answer – we’re just enjoying each other’s company. But what are we supposed to do instead? Read more, then chat? Plainly, yes; and yet AI would be faster, being as it can read everything, all at once. That’s why I wouldn’t want to discuss it at a party; as for why Prof Computers wouldn’t – well, I’d have had to have asked, which would have been yet more (rude) chat.
The sister of the birthday person gave a beautiful speech, in which she quoted what someone had said at their mother’s 80th: “She walks, she swims, she drives, she lives!”
ChatGPT cannot get better than us at doing those things.
• Zoe Williams is a Guardian columnist
Do you have an opinion on the issues raised in this article? If you would like to submit a response of up to 300 words by email to be considered for publication in our letters section, please click here.
