Is artificial intelligence lining up to be the perfect scapegoat? That’s what Christopher Nolan seems to think, having told Wired in a new interview published today that the biggest problem he sees with AI is how it’s being framed as an omnipotent resource, making it a great and plausible thing for companies and others to blame their mistakes on. “I identify the danger as the abdication of responsibility,” Nolan explained. Elsewhere in the interview, the 52-year-old director, whose atomic bomb film Oppenheimer opens July 21, revealed that he was going to die for sure in a nuclear holocaust.
From a Wired feature:
Nolan: […] Everybody has a very—call it a partisan point of view. The issue with AI, to me, is a very simple one. It’s like the term algorithm. We watch companies use algorithms, and now AI, as a means of evading responsibility for their actions.
Wired: Say more about that.
Nolan: If we endorse the view that AI is all-powerful, we are endorsing the view that it can alleviate people of responsibility for their actions—militarily, socioeconomically, whatever. The biggest danger of AI is that we attribute these godlike characteristics to it and therefore let ourselves off the hook. I don’t know what the mythological underpinnings of this are, but throughout history there’s this tendency of human beings to create false idols, to mold something in our own image and then say we’ve got godlike powers because we did that.
Wired: That feels very, very right now. Like we’re at that tipping point.
Nolan: Exactly.
Wired: With these large language models, the machines might even be able to teach themselves the next step.
Nolan: There was an interesting article in the LA Times about ChatGPT and OpenAI. It basically said it’s a sales pitch, that they’re a private company now. And they have the greatest sales pitch in the world, which is, This is a really dangerous thing. Maybe we shouldn’t put it out there. So now everyone wants it. Doesn’t mean there isn’t a real danger here, because I feel that there is. But I personally, and this is just my opinion, I identify the danger as the abdication of responsibility.
Discussion (8 replies)
Join Discussion →Chris getting a little melodramatic there at "old age"?
Blaming ai is like blaming CGI, it is not the technology that is bad, it is the idiots over relying on it and using it for everything.
100℅ ai is but a tool, a good tool. When you have people talking about solving this, and that all kind of human problems and making better beer, well that is the problem. You assign it definitive authority, and then shortly you run into problems.
Yes, AI has no critical thinking skills, it is an aggregator or knowledge, it is up to the user to check and verify the accuracy of the output.
Even if you have to iterate a dozen times before it gives usable output, it is still 100x faster than doing the work manually.
Ai today is only as good as it's source data. Open ai platforms regularly give crap responses that seem valod but are vapid at s minimum.
This is why there are projects at MIT to validate data sources and grade them for AI consumption.
Mr Nolan has a point and the uninformed will blame AI and not validate information.
This is why it's critical for people to be able to have ai that digests trusted data sources and not any. A single AI for every potential data source is a neat expirament to play with but not something that should be trusted.
I think Nolan is a genius
Now I don’t have to blame the dog for midnight farts in bed, I can blame AI
What are the odds someone is working on an AI to be used to validate the data?
I know you're joking but it's already in the works. I did a story about a month ago where a researcher on AI said he'd already heard the next Gen AI was already planned to start overseeing its own data.
Midnight farts? Maybe.
Fluffing the covers afterward? That's all on you, buddy.