When it comes to artificial intelligence, or AI, people tend to fall into two camps. One camp, best represented by Mark Zuckerberg, believes that AI is the way of the future — that we will use little programs to ease our lives and make us more productive and happy. The other camp, best represented by Elon Musk, is that the machines will become smarter than we are and possibly kill us all. And now Musk and Zuckerberg are having a catty, hilarious nerd fight about it. And the best part is that they’re both probably wrong.
It all started Sunday, as all ridiculous arguments between celebrities do, on social media. Zuckerberg asked about Musk’s stance on AI, calling it “irresponsible.”:
With AI especially, I’m really optimistic, and I think that people who are naysayers and try to drum up these doomsday scenarios … I don’t understand it. It’s really negative, and in some ways, I actually think it’s pretty irresponsible.
This morning, Musk responded with the nuclear weapon of nerd arguments, namely insulting Zuckerberg’s intellect:
I've talked to Mark about this. His understanding of the subject is limited.
— Elon Musk (@elonmusk) July 25, 2017
What’s kind of funny here is that Musk is the guy building commercially available robots with crude AI that, arguably, are already killing humans, albeit accidentally. At the same time, considering the disaster that has been Facebook’s 2016 — the site was turned into a propaganda-spreading factory — Zuckerberg’s optimism seems a bit misplaced. So who’s right, in this debate? Neither, really.
First of all, it’s rapidly becoming clear we don’t know the limits of machine learning and algorithms because, in ways both horrific and hilarious, we’re slamming right into them. Machines, unsurprisingly to everyone but Silicon Valley types, don’t interact with data or look at the world the way we do, and the future of AI research may lie in coaxing robots to better understand the world from a human perspective.
Secondly, the future Musk is panicking about, to the point that he’s trying to build cyborgs, seems increasingly unlikely to happen. When AI have gone up against humans, usually it hasn’t gone well for the AI. Introduced to the internet, Microsoft’s “Milllennial chatbot” Tay was turned in a screaming racist mess in hours. Facebook’s own vaunted algorithms have turned out to be a lot crappier, unsurprisingly, than Facebook is willing to admit.
Perhaps the most important limit to artificial intelligence, in the end, is the humans who program it. The tools we create reflect our follies as much as they reflect our intelligence, and any AI is ultimately at the whim of the humans who build it, whether those humans realize it or not. We should be more worried about the people behind the AI, not the AI itself. But, hey, if all Musk and Zuckerberg want to do is argue like schoolkids, that’d be a good use for AI. In fact if we could build AIs to just have pointless arguments on the internet for us, that would save an awful lot of time.
(Via Business Insider)