If you’ve been following the debate over generative AI, you probably know that the conversation tends to break down into two camps: evangelists and naysayers.
You can find just as many people who think this technology is going to save the world as you can folks who believe it will bring about the end of civilization as we know it.
Writer Paul Ford has been covering technology for decades, and in moments like this, he has a mantra: “It’s just software.” If you can separate yourself from the heated debate, Ford says, you can see technological innovations for what they really are.
And as he writes in his latest piece for Wired magazine, in the case of generative AI, there’s nothing to fear — except, perhaps, ourselves.
Full conversation
PAUL FORD: It is a robot, and it has no idea what it's saying. It really doesn't. It seems like it does. But it doesn't. It is just wacky Mad Libs going at a mile a minute, taking all the data that's in the world and aligning it with what you jammed into the box and saying like I think you want me to say this, but not even that. It's not even — it's a talking spreadsheet.
And it's really fun, because — I don't know about you. I mean, you work at a NPR affiliate, so you are probably extremely driven by shame — and like…
[LAUGHTER]
SAM DINGMAN: No comment
FORD: Yeah, I mean, I'm a writer. And I'm always editing in my head what I get to say, and then the robots showed up like a bunch of drunken uncles just rambling about absolute nonsense. It's like a mixture of a prize salesman and a goofy toddler. And it's really ridiculous.
DINGMAN: Yeah, yeah. It feels to me a little bit like what you're saying is that we're so busy endowing AI with all of this power — whether we think that power will be used for good or for ill — that we have forgotten to pause and realize, to use your word, how goofy it is.
FORD: Look, it does a magic trick. Eventually, we're all going to kind of know how that magic trick looks and you're going to — who goes back and sees a magic show the second time, right? You can now — everybody's increasingly, like — you can see the weird edges on the images that it generates. It's not just that it puts in multiple fingers. You can tell something feels off.
And so, we're gonna get an immune system — a cultural and mental immune system — and we're going to just get used to it, and then we'll find the things it's actually good at.
DINGMAN: I feel like what you're expressing is, in a way, quite profound because, thought about in these terms, it is very easy and perhaps even advisable to dismiss AI to a certain degree as a kind of an unserious thing that we are in a rush to make very serious.
But there is also this way in which it does represent something very real about human psychology, which is that we all kind of have this desire to be like, is this what you want? Is this what you want? How about this? I'll keep trying, I'll keep trying, I can do better. And there is something weirdly relatable about that.
FORD: Well, I mean, remember "Castaway?" I still feel sad for that volleyball. It did its best. We anthropomorphize absolutely everything, from stuffed animals to mascots at the sports team. We anthropomorphize everything.
And I think what is tricky — I remember there's a very good thinker, a very practical thinker on this, a guy named Simon Willison. And he sort of feels that the cardinal sin of this world is the anthropomorphization — like making them seem like people, giving them names, constantly trying to fake a human on the other side as an assistant as opposed to just producing the data in a more overtly robotic way. Because you're hacking human psychology when you do that.
You're not actually adding a lot of value or teaching people a lot about the technology. You're teaching them that "yay, really treat it like a human and it'll work out." And that actually doesn't get people into what these things can really do. They can do all sorts of wacky stuff that maybe if you weren't thinking of it as a human, you might be more likely to explore.
DINGMAN: Right. Well, because in a way, generative AI is able to do things that most of us are not able to do as you write, "It does everything badly and confidently, and I want to be it."
FORD: Well, it's that shame thing again. I mean, imagine just — you see how much fun it is to be shameless, right? We have political candidates who are utterly shameless. They just get away with, they say whatever they want, and they get away with it. And it can be a little horrifying, but oh my God, imagine just being that free for like five minutes, just making stuff up. I just feel, it feels great. They don't ask permission. It's all forgiveness.
DINGMAN: Speaking of the real-life analogs of the this type of shamelessness that we see AI reflecting back at us, one of the comparisons — or alignments, perhaps would be a better word — that you do make in the piece, is that the way that AI behaves sort of mirrors the, at least, public behaviors of the people who are constantly out evangelizing it, constantly out saying the world wants this technology that so many people are saying back to them: "We did not ask for this. We don't want it and, by the way, we're suing you because you stole all our work." And yet, they just keep going, saying like, "No, no, this will be the future."
FORD: If you want to go meta, like really, really meta for a minute, right? Like, one of the things I'm talking about in the piece — but I think this is real — there is no superpower like shamelessness. You can get to incredible heights in our culture.
And so, you have people running AI companies saying this is actually going to become intelligent really soon. There isn't a clear path. There isn't a way to get there that I see. But they're just promising. Like, it's just gonna keep getting better, faster and faster, just you wait. But the ability to just get out there and just throw it out there and be like, "Hey, you guys figure it out. I'm gonna tell you what you should believe," is a superpower.
And people who footnote everything and have to get fact-checked when they write their magazine articles, they don't have that superpower, right? Like they're accountable in a different way. And we like that. We like that in our society. We say we do.
But then we go to the concerts and vote for the people who are the most shameless. And these robots have it in spades.
DINGMAN: Yeah. But it does seem to me — if I'm reading your piece correctly — that when the tool is, perhaps inevitably, improved, something — in your mind — will be lost.
You write that these tools as they currently exist, "mirror the arrogant, repetitive ramblings of our betters. The horrific confidence that keeps us driving over the same cliffs. That arrogance will be sculpted down and smoothed over. But it will have been the most accurate representation of who we truly are to exist so far. A real mirror of our folly."
So is that fair to say? You think, in a way, we will lose something when generative AI gets more useful.
FORD: We always lose when the first days are over, right? The fantasy has to go away.
The early days of the web, we were going to publish our zines online and use things like GeoCities and there would be animated GIFs and we would all connect as a culture and it would just be the absolute best future possible and everyone would like publish their fan fiction. It was really, you know, it was full of light. And now it's for banking and harassment.
So it's sort of like it's the same thing. We're in the zone of folly. You can say absolutely anything about this technology, and no one can really prove you wrong. And, at the same time, because they've tried to make this thing seem like a human, it's the worst human.
It's just absolutely terrible at everything it does, but the fact that it can do it - it's the joke about, "but such large portions!" Like you can just have so much of it, that it just seems amazing and ridiculous all at once.
DINGMAN: Kind of like humanity, I suppose.
FORD: One can hope. That's probably the best we're gonna get to, right? Let's hope for amazing and ridiculous, and I think that's us at our best.
DINGMAN: Well, Paul Ford is a writer on technology and the web and many other things. In this case, he wrote "Generative AI is Totally Shameless. I Want to Be It" for Wired Magazine. Paul, thank you so much.
FORD: It was my privilege. Thank you.
