Whatâs behind the AI industryâs latest warnings of doom?
The AI industry seems to be having its loudest debate yet about whether its technology poses an existential threat to humanity. The current discussion began afterAI researcher Jacob Coxon said that heâs resigned from Anthropicbecause heâs worried that the leading AI companies are âgambling with our lives.â Then Anthropicâs alignment leaned chimed in witha post declaring, âWe really do earnestly believe AI could kill all humans!â adding that he personally thinks the chance is â>10% within the next decade.â On the latest episode ofTechCrunchâs Equity podcast, Kirsten Korosec, Sean OâKane, and I discussed the latest apocalyptic warnings. I tried to articulate why Iâm skeptical of many AI doomer narratives, while Kirsten asked if this was âjust a weird way of flexing to show how far advanced their companyâs AI model is,â particularly as these companies prepare to go public. And Sean wondered how these concerns might show up in Anthropicâs S-1 filing for its IPO: âAre there junior lawyers right now who are going through and having to rewrite that entire section of the S-1 filing to say, âItâs a officially Anthropicâs position that thereâs a more than 10% chance that we could develop something that would eradicate all of humanity and that would be materially bad for our businessâ?â Keep reading for a preview of our conversation, edited for length and clarity. (Note: We recorded this episode before Anthropic CEO Dario Amodeipublished his plan for more cautious AI development.) Sean OâKane:Iâm hard-pressed to think of something that blew up so fast. Not only did this warning shot come out from this young researcher who has also worked at OpenAI, but also was immediately shared on X by the alignment lead at Anthropic â who, in what might go down as one of the best misplaced exclamation marks ever, shared Coxonâs post and and thread and said, âWe really do earnestly believe AI could kill all humans!â Exclamation mark! What a weird vibe. That was just a ton of accelerant on an already fraught post or series of posts. Coming after the Hugging Face hack from OpenAIâs internal model, plus just the increased capabilities weâve seen with the latest models released by Anthropic and and now OpenAI with with Astra a few weeks ago, I think this was just perfectly timed to be a powder keg type of thing for this young researcher to say. Anthony Ha:Just to disagree with you, I do think that if you believe that AI could destroy all humanity, that does deserve an exclamation point. I would argue that that is a perfectly well used exclamation point! My issue with that tweet was more the âwe.â Who is the âweâ here? To what extent can we talk about sort of the AI community or AI research community as a monolith? And the greater than 10% chance â thatâs just a made up number, that doesnât mean anything. Sometimes [there is] this habit in both the tech industry and other places to just throw out these percentages, theyâre not based on anything or calculated based on anything. [In retrospect, I realize the tweet was probably referencingthe concept of P(doom), but I still think itâs silly.] One thing I will say about Coxonâs statement and decision is â thereâs this recurring theme on Equity, when someone like Sam Altman or Dario Amodei is doing this doomer narrative, thereâs always this element of: Well, then, why are you doing what youâre doing? If you actually believe that [AI could destroy humanity], you would not continue doing this. [Whereas] this is actually somebody putting his professional trajectory where his mouth is. Heâs actually saying, âI believe this is really, really, really bad, and I donât want to keep working on it.â And so, props for having the courage to do that, if nothing else. Kirsten Korosec:Yeah, I put him in a separate camp than everyone else saying that and talking about the dangers. Iâm going to put my speculative hat on, because I want to ask both of you a question, which is: Is it possible that every single time we see the increasing number of blog posts about yet another incident in which one of their AI agents breaks through unintentionally, or they talk about how humanity is at risk, is this a weird way of flexing to show how far advanced their companyâs AI model is? I mean, that sounds very cynical, but it does achieve that purpose. Which is: If these AI models werenât advanced and werenât capable and werenât breaking through, we wouldnât have to worry about these things, right? Itâs like a very weird way to brag about the capabilities of the models that youâve created within your own company. Anthony:Iâve definitely wondered about this. I donât think itâs completely cynical, in the sense that I donât think itâs all just a very conscious marketing ploy across the board. I think that when a lot of these people â whether the researchers or CEOs â talk about it, they do have real concern. But of course, it does align with [their] business interests in a lot of ways, to say, âWow, weâve built the most deadly software thatâs ever been made.â I donât want to get too psychoanalytic here, but others have pointed out that there is this temptation on a personal level of: Of course, you want to believe that the thing youâre working on is the most important and most dangerous thing in the world. Sean:The thing that sticks out in my mind when I think about that question is, thereâs certainly an element that makes it seem like, âOkay, weâre doing this thing thatâs so capable, and thatâs good for us in some way, even if it looks bad in a lot of different lights.â I think whatâs different about some of these most recent examples is, it really gives you the feeling that these companies donât have a handle on this stuff in certain ways, especially with the OpenAI stuff. We keep seeing more and more reporting about otherinternal agents that have accessed different wikis on the weband are leaving messages for each other, and in a way that doesnât seem like itâs being handled in a competent way from OpenAI. I would imagine there would be just a bit more polish on the story being told, if it was wholly about getting people to believe that, âOh my gosh, theyâve made something so incredibly capable.â The other thing that I think is really fascinating about this, in particular, [is] weâre what, a few weeks at most out from seeing Anthropicâs S-1 filing for its IPO, and just a couple more weeks or month or two away from a potential IPO. And the idea that youâre going to come out and say these things in this clear language ahead of an IPO â Iâm very interested in what that means for that process. How much of this kind of stuff had they already written into the S-1 and the risk factors inside that document? Are there junior lawyers right now who are going through and having to rewrite that entire section of the S-1 filing to say, âItâs a officially Anthropicâs position that thereâs a more than 10% chance that we could develop something that would eradicate all of humanity and that would be materially bad for our businessâ? Kirsten:Youâre assuming that itâs not in there already. Sean:Thatâs what Iâm saying, though: Is it in there already and being reworded? Or is this something thatâs a true scramble? There has to have been language in there. Itâs one of the reasons Iâm so eager to read this document in a way that goes even further, in some ways, than the SpaceX [S-1], because Iâm sure that thereâs probably stuff specific to these ideas that will be interesting to see. Kirsten:Hereâs the thing: In a traditional investment environment, one might believe that language like this would hurt the valuation of a company, because itâs suddenly dangerous. But we donât live in normal times. And so again, back to my point, it could end up being a weird beneficial flex for the company on the valuation side. Itâs not the same as the whole rage-baiting trend that we saw last year, but itâs in that same, letâs say, universe, in which the strength, capability, even elements of danger of something, equals high valuation. So I guess weâll see in a few weeks. Putting that aside for a minute, what is being done about it? And can we control this? Tthe U.S. executive director of a nonprofit called ControlAI, Connor Leahy,he was on the show this week, talking about this. So what are you paying attention to in terms of how to control the dangerous aspects of AI, or are we throwing up our hands and watching it all unfold? Anthony:I myself do not necessarily have a great answer to this, but I have been thinking about some aspects of this debate and maybe why I respond the way I do. To echo one of Seanâs points, I do think that part of what this speaks to is the extent to which these major AI companies are feeling like theyâre not really in control of these models anymore. Thatâs definitely not great. That is something that we should all be worried about. I do think that part of the reason Iâm skeptical of the doomer narrative or resistant to the doomer narrative is because it reaches this level of hysteria of, âWow, this could destroy humanity in the next 10 years.â It is a little bit of a distraction from the more immediate harms that AI can have, whether thatâs labor-related, whether thatâs environment- and climate-related. Ideally, I think we should be able to discuss all of these things, and have regulatory and other kinds of safeguards against all of these things [including AIâs existential threat]. But once you start using phrases like AGI and superintelligence, that just sucks up all the oxygen in the room in a way that is not very helpful.
