AI Judge changed title after evaluation, original title: "Andrew Mayne, formerly of OpenAI, argues human trust networks cannot be automated, countering economic models of total labor displacement"
Andrew Mayne, who spent four years at OpenAI working on GPT-3 through GPT-4 materials and now hosts the company podcast, pushes back on the premise that falling compute costs will let AGI replace every economically relevant human task. His October 27 essay targets Pascual Restrepo’s draft paper by stressing that trust, relatability, and preference for human interaction sit outside pure task economics and resist full automation.
@deanwball And it's absurd to think I could be wrong.
Lord, grant me the cajones to one day begin my essay: “Let’s start from this premise: I am correct. My conception of cognition, AI, labor markets, the economy, human beings, and the world is perfect, and things will go exactly as I expect. I can’t prove this, but…”
Mayne notes that therapy, physical therapy, live events, and brand experiences often stay human-led because audiences value authenticity and peer networks over benchmark wins, leaving open how much of the economy actually follows cost curves alone.
The piece recalls that past machinery shifts raised demand for roles built on human connection rather than causing permanent mass unemployment, though it leaves unclear whether AGI timelines change that pattern.
Many users insulted defenders of the AI-can-replace-all-workers essay with accusations of bad motives and lack of real jobs, while some endorsed reading Ricardo first for historical context.
No Digg Deeper questions have been answered for this story yet.
Dean has sizable cajones himself if his poasting game is any indication; but this response to @zetalyrae is weak. Let's start with the fact: "we" (not me sadly) are explicitly raising trillions on the pitch of "a country of geniuses in a datacenter by 2028". How do you plead?
Lord, grant me the cajones to one day begin my essay: “Let’s start from this premise: I am correct. My conception of cognition, AI, labor markets, the economy, human beings, and the world is perfect, and things will go exactly as I expect. I can’t prove this, but…”
@deanwball I think it’s a stronger premise and essay than this thread suggests but also he did take some unfair potshots at you so I get it
Lord, grant me the cajones to one day begin my essay: “Let’s start from this premise: I am correct. My conception of cognition, AI, labor markets, the economy, human beings, and the world is perfect, and things will go exactly as I expect. I can’t prove this, but…”
@deanwball What does "superintelligence" (which the big labs all assert they are working on) mean if doesn't mean greater than human ability at every task? If it doesn't mean that, then it's just a glorified estimator of Internet n-grams, no? Btw it's spelled cojones.
@alexolegimas an essay going around called "nobody escapes the permanent underclass" which for some reason is gaining traction despite is being a rehash of a type of essay that has existed since, depending how much you're willing to squint, somewhere between the mid-1840s to 2014
lord, grant me the cajones to one day begin my question: "how would you feel if you hadn’t eaten breakfast, i am correct. i know you didn't breakfast even tho you might have. I can't prove this, but..."
bro learn to read
Lord, grant me the cajones to one day begin my essay: “Let’s start from this premise: I am correct. My conception of cognition, AI, labor markets, the economy, human beings, and the world is perfect, and things will go exactly as I expect. I can’t prove this, but…”
@deanwball idgi he says he wants to use this as a premise and he can’t prove it. what about this is undeservedly bold?
"Let's start from this premise: I am a genius. Everyone must bow down to my towering intellect. I am right about everything. I can't prove this, but the goal of this post is to argue that *if* I am right, then everything else follows."
I am being serious about cajones though. in college we'd have identified this trait as "spittin' game." "let's start from the premise that I'm right" is exactly the energy required to walk into a bar and get laid.
Link to the whole thing: https://www.marxists.org/reference/subject/economics/ricardo/tax/ch31.htm
@deanwball Your sneering does you no service, as many people who believe the first sentence go on to believe in the permanent underclass. Can an essayist not argue for “A->B” anymore without inviting ridicule?
@deanwball I mean, I read this more as just being upfront that this assumption (AI is a direct substitute for human labor) was being made, not necessarily that it was true (or even necessarily very likely). Could probably have been better worded, though!
@deanwball Ah, yes, the "all labor will be automated by assumption" essay. Earliest instance I've found was 1817, On Machinery by David Ricardo (who had some very good ideas as well, tbc).
@deanwball @zetalyrae Congrats. More seriously, what part of "workers will be replaced" does not follow from AI being Pareto-superior to them? Hard mode: don't give me "comparative advantage".
@teortaxesTex @zetalyrae guilty to being in an industry that I expect will profoundly better the human condition
@shakoistsLog @deanwball You need a very high IQ to understand the difference between P ⊢ Q and ⊢ Q.
hmm that's very much not what he is saying
he is addressing a very specific claim by others that IF his premise was true, some people might escape it through capital or tribal means
i'm not the only one reading it this way
perhaps i'm missing some nuance you're picking up or assuming?
@alth0u correct. he is assuming his conclusion, which is that we will all be replaced by ai because ai will do all of our work.
@deanwball where is this from....
That's largely fair but a bit handwavy. I am a proponent of scenario modeling, and a scenario should not be dismissed with an appeal to "cajones". My issue with comparative advantage is that a) it doesn't put a floor on compensation for the potential Pareto-dominated employee, and b) processes have costs beyond labor, so employing high-error-rate, slower laborers no matter how cheap can be prohibitively costly for an enterprise. By the same token, we don't employ obsolete LLMs. Fernando's thesis is deliberately provocative. I disagree with him myself, though only on stages 2 and 3 (replacement of the elites and the state); as for workers – all or as close to all as to be the same thing socially speaking – I struggle to argue that this doesn't have a very high likelihood of happening. It would be nice if AGI labs had some coherent retort.
of course workers will be replaced. perhaps a great many of them! "all" however, is an awfully big ratio, and assumes away the most interesting part of this discussion in the name of writing a piece with a compelling title. what I would say is that tasks and production processes do not exist a priori of production; production is itself a dynamic process where the input tasks are constantly being discovered and refined. thus I would expect, in the long run, for humans, firms, and the economy to be dynamic in ways that are hard to predict but ultimately produce jobs that humans can do. some of this will be because of simple or complex human preferences, some of this will be because of your apparent villain 'comparative advantage,' and some of this will be not quite because of either of those things but because of the intrinsic dynamism of people and of the world. now, this does not mean things will go great! the agricultural revolution probably sucked for many, I'm with James Scott on that one. ai could reconfigure the world in a way that is bad for many people, but assuming the conclusion of "all" negates precisely this more interesting and complex debate.
@eznark @alexolegimas yep. "here's my conclusion, I'll present it upfront so I seem like a clear writer, but I'm actually putting the conclusion at the front so I can assume it. by the way, you cannot definitively prove my conclusumption is wrong"
@shakoistsLog a)the claim "it's absurd to say you can disprove this" is just a truism, ofc you can't disprove it, so this falls into the category of sophistry for me, and b)you're hiding away all the interesting part of the discussion by assuming your conclusion!
@teortaxesTex @zetalyrae guilty to being in an industry that I expect will profoundly better the human condition
