Adam Conover's Conspiracy Theory Doesn't Make Sense

I like Adam Conover.  I've watched him since the CollegeHumor days.  He's funny, he's smart, and I agree with a lot of the stuff he's said in the past.  I'll probably agree with a lot of stuff he says in the future.

But recently, he uploaded a video, "AI Existential Risk Was Invented to Distract Us" which seems pretty half-baked.  

Now I don't really think that it's very likely that AI will kill everyone.  I think it's much more likely that AI enslave us or make us doglike.  (To see my list of probabilities, click here.)  That said, I don't think it's impossible that AI will kill or hurt a lot of people.  Just unlikely.  And even though it's unlikely, it's important that we take steps to make sure it doesn't happen.

But Conover thinks that people like me, who are concerned about this, are just big dummies, suckers, rubes, who have had the wool pulled over our eyes.  According to Conover, we are victims of a conspiracy - a very complicated conspiracy, one even more complicated than the one in the conspiracy theory that Donald Trump has been promoting.

I don't believe in the conspiracy theory that people like Donald Trump and Curtis Yarvin are promoting, for at least 3 reasons:

1. I try not to believe in conspiracy theories at all, no matter who is advocating them.  I'm more skeptical and apply greater scrutiny towards conspiracy theories than I do other kinds of explanations, because I know that people have a tendency to come up with conspiracy theories, and they tend to have a strong psychological grip on a mind - once they're in there, they can be very difficult to dislodge.   

2. As it happens, Donald Trump and Curtis Yarvin are my political enemies.

3. When you scrutinize it, there's no evidence for this particular conspiracy theory, and it doesn't make sense.

But Adam Conover's conspiracy theory is even less plausible than Trump & Yarvin's conspiracy theory.  According to Trump and Yarvin, AI is harmless and indeed profoundly beneficial to humanity.  They seem to be hinting that China is behind the AI safety movement - supposedly, China is doing a psyop to hold back the development of AI in the USA so that China can catch up and pull ahead in the AI race.  At a superficial level, that at least sounds like a plausible story, but when you dig into the facts, it doesn't hold up.  China doesn't seem to be concentrating that many resources towards AI.  They're not all in on the AI race.  And they really have no hope of catching up any time soon.  The USA has far more datacenters than all the other major players combined.  China has relatively few, especially per capita.  Even in absolute numbers, France has more datacenters than China.  So much for the Trump theory.

But Adam Conover has a different conspiracy theory.  According to Conover, AI really is harmful, but AI companies are trying to cover up the real harms of AI by making it look even more harmful.  What?  That doesn't even make sense at a superficial level.

His argument seems to be: Seems like a lot of the people involved in developing frontier AI - the engineers and the CEOs and a lot of people who work for these companies - are concerned about AI safety.  Seems a little suspicious, if you ask me.  If all these experts are concerned, how come I'm not concerned?  How come I'm just learning about this?  If I didn't grow up thinking about this, then it's probably not real.

Imagine this kind of argument applied to any other industry.  "Seems like a lot of people who run nuclear reactors are concerned about the possibility of a meltdown.  It's almost like they want to prevent that from happening, like it's their job to prevent it from happening.  But they work there.  Seems fishy."

Does he have any evidence for this conspiracy?  If so, he doesn't present any.  It's all just vibes.  "This feels true to me."  You go, Adam.  Speak your truth. 

So for instance, when Conover considers the possibility of an AI developing a bioweapon in a lab, he says:

"We could all just... not do that.  Right?  We could all just choose not to build the lab.  In fact, if we were to build such a lab, would it be accurate to say that the AI designed the bioweapon?  Or would you say that we, the humans, designed the bioweapon by creating the exact, like, necessary conditions for it to happen?"

Okay, a couple things.  

First: I don't care.

This seems like a purely semantic distinction, without any real difference.  I don't care whether you think it's technically "accurate" to say that AI is going to do something that kills a lot of people or whether you think it would be better to say that humans are going to do something that kills a lot of people, using AI.  These sound like the same thing to me, more or less.  Or whatever.  I don't care.  This seems like pedantic squabbling - the issue is the harm that could come to people, not the precise language we use to describe that harm.  Yes, okay, fine, Adam, let's go with your language - whatever it might be.  I don't want that.  I don't want humans using AI in a way that could cause a lot of harm to a lot of people.  So shut up with your language policing.  

(Similarly, he says that the Hugging Face hack shouldn't be considered an act by the AI's "own volition," but that effectively OpenAI employees did it ("You are Pagliacci, motherfucker").  Again, okay, fine, whatever.  According to Conover, the hacking incident happened, but doesn't really matter.  Instead, "the question is 'How do you frame what happened?'"  Again, enough of this semantic bullshit.  I don't care.  Frame it however you want.  It sucks, and we should stop it from happening again.)

Second: "We could just... not do that. Right?"  Right.  So let's not do that.  That's what we're saying.  But also remember the immortal words of "squareallworthy":

Everyone will not just If your solution to some problem relies on "If ... 

 

This applies here.  We cannot trust that people will never use AI in such a way that it would harm people.  It's going to take legislation to prevent it.  In fact, it's going to take international agreements.

Oh and by the way: AI biolabs are not some fictional thing that could exist in a hypothetical future.  They already exist.  The labs have been built.  Here's one.  Here are 19 more.  Andrew Ho of OpenAI is building his own.

Comments

Post a Comment

Popular posts from this blog

Lenin the Moralist

  "Unprincipled practical politics are the most impractical politics."  - V. I. Lenin  It will probably cause me to lose friends all across the political spectrum when I say that the problem with Lenin was that he was too moral.  Deal with it.  Moreover, modern-day soi-disant Leninists are even more moralistic than he was - and therefore, all the more romantic and delusional, and their analyses all the more pointless. I do not mean to imply that Lenin was not responsible for the deaths of many people - he was.  Nor am I saying that Lenin did not commit enough atrocities - that he was not violent enough, not ruthless enough, too timid to accomplish his goals, or anything of the like.  No, no, no.  My point is that Lenin's ideology diverted sharply from Marx's critique of political economy, and in part this was because he tried to reinterpret Marxism into a kind of moral campaign, when Marx had intended nothing of the kind - and, by the way, Lenin...
Read more

Why Capitalism is Ending

Image
  Sorry, Ray Kurzweil , there will be no singularity.  As much as Moore’s Law has become a cliche, it has also become a cliche to point out that exponential growth has no "knee" - that is to say, that an exponential growth curve has no inflection point.  It goes up faster and faster, so that not only is the value rising but the derivative of the value is also rising (that is, the rate of increase is itself increasing) yet at any given point it is still a gradual increase.  Growth that gets indefinitely huger as it approaches a specific point of time, a limit known as a singularity, is not exponential growth but rather hyperbolic growth .  There’s no evidence that technology is growing hyperbolically; it is “only” growing exponentially.  Gordon Moore himself has stated that he doesn't believe in the singularity, or even in the continuation of Moore's Law.   (And Moore's Law is starting to fail, anyway ... )  But it hardly matters;...
Read more
Image
The right is convinced that politics (and economics) are downstream from culture. But the right is winning on the economic front, while losing on the cultural front. The left maintains that the economic base of society in the final instance determines its cultural (and political) superstructure.  ("It's the economy, stupid!")  But the left is winning on the cultural front, while losing on the economic front.  In other words, both sides see themselves as losing in the way that matters to them. Both sides see themselves as winning only in a superficial and ultimately even counterproductive way, while losing on the front that is fundamental and primary.  How can this be true? Why aren’t both sides winning on the issues that truly matter to them, especially since the opposing side doesn’t really care about those issues? Why aren’t we living in an egalitarian society in which class and other forms of inequality have been abolished, which is at the same time a so...
Read more