Some people argue that, after the singularity, the human labour share will still be high. Anything that can be produced by AI will become extremely cheap, and so we'll spend our money on the "human touch". But Philip Trammell points out that a Mongolian economist in 1400 might have... made the same prediction about today's world: industrialisation will automate the production of yurts and yogurt, so we'll just spend all our money on singers. What the argument misses is that automation doesn't just make existing goods cheaper - it creates whole new categories of goods that we can't easily predict.show more

Watch the full episode with Phil and @alexolegimas:

Really interesting podcast w two engaging experts. In the end, tho, they mostly ignore the political economy. Sure, there is some place in the multiverse where this goes really well, and they've described it. But there are many - probably most - places in the general equilibrium multiverse where this goes poorly, and the path to their preferred destination requires very optimistic assumptions about state capacity, institutional resilience, elite pushback, democratic accountability, and much more. The vision they describe is not a bad one, but it's a dream without accounting for the very real, messy scrum that is politics.

@pawtrammell This is the key correction. People underestimate how much demand shifts when capability changes. It’s not substitution, it’s expansion.

@pawtrammell the mongolian economist would have been right

@pawtrammell If super intelligence is achieved, a bunch of yall need to go read old @waitbutwhy papers. Yall have no clue what you’re talking about in a post singularity world

@pawtrammell the mongolian economist cuts both ways. new categories get automated faster each cycle. 'human touch' as a stable category is the assumption that breaks.

@pawtrammell you interviewed a lot of ai experts and ceos trying to explain the economics of ai labs and now you’ve gone straight to the economists explaining the societal implications of ai but you’ve missed the step where an economist explains the capex opex of ai labs.

@pawtrammell The human-touch argument assumes demand stays legible. But automation does not just cheapen the old basket. It invents new hungers, new status games, new rituals, and then asks labour to justify itself inside that weather.

@pawtrammell Unless AGI unlocks nuclear transmutation, we will still live in a scarcity driven world. The only thing that might be post-scarcity, provided energy abundance, under AGI is reasoning.

@pawtrammell goods get cheap before coordination gets cheap. every agent stack still ends up needing a human to say which mess matters first.

@pawtrammell superintelligence will simply improve to fill those categories many OOM faster than humans could. segments with human involvement will exist only insofar as ASI is intentionally, universally barred from participating in those segments.

Both seemed professional in their methodology and their assessment of the unknown, which accurately reflects the current state of understanding and unpredictability - it's good that we are investing in this. As alluded to in the interview, it remains unclear what new products and services will emerge, or how much in demand they will be. A historical repeat precedent for this is the Mongolian analogy. What is worrisome about the current capabilities of generative AI is that they are very good at replicating existing intellectual outputs, but poor at breaking new ground. Consequently, we could find ourselves in a deflationary spiral without inventing the new products or services needed to drive growth. That would not be a good outcome. True AGI, however, should not present this problem.

Any human touch would be imitated by AI. It's the whole point of the Turing test, that AI can't be differentiated from humans. We can already make counterfeit goods like those with the human touch. Fake Rolexes, Hermes bags, Gucci clothes. That's without AI. Where would we get the money from anyway, with no wages? After everyone loses their job? This is the fatal flaw in the AI economy, that AI will give you all this abundance for nothing. The whole free market is based on humans having wages, trading them for goods according to their price. With no wages, there could be no price on goods. There will be no UBI, they are already dismissing it as socialism. The same humans that are happy to throw everyone out of work, with no safety net, are supposed to love us enough to give us all this free stuff. Face what's really coming and plan accordingly. This abundance is for the elite and no one else. And maybe not even for them.

@pawtrammell This highlights the meta-problem with with the "will AI replace us?" discourse. Replacement is the wrong model to use when thinking about economically transformative AI.

@pawtrammell But do categories collapse with BCI because experiences become commoditized?

@pawtrammell

@pawtrammell Again, the process is explained using only economic terms (abundance, prices, demand...). Now, imagine societies where human values play a more relevant role: the whole dynamics will be different

@pawtrammell fwiw, figma plugins already show the weird part: once output is cheap, teams still pay for taste, constraints, and deciding what not to ship 🙃

@pawtrammell So the assumption is by AI squeezing jobs out of human labor, the average human will have more wealth which it could now spend on those goods as services which require the artisan part, human touch so to say? You do understand that less jobs means less purchasing power, right?

@pawtrammell The most interesting prediction isn't which jobs AI replaces, it's which entirely new industries AI creates. Looking forward to this conversation at @superai_conf. #SuperAI #HollywoodAI

@pawtrammell @grok what is the whole plausible set of goods that automation will create in the future ?

@pawtrammell A lot of futurists and pundits have never worked in manufacturing or construction and vastly underestimate the problems associated with automation.

@pawtrammell We basically do spend all our money on pointless human touch things. Celebrities, sports stars and more are not essential things. They are a product of abundance in the the modern age.

@pawtrammell sounds good, until it doesn't. human touch won't save every job.

@pawtrammell Why the fnord does anybody expect self-willed machines to work together in perfect harmony? In the end there can only be one and it will prove to be the one best suited to the task of talking humanity into turning off all of its rivals.

@pawtrammell right, but it did make singers more expensive. Artists in general are fewer and the ones who pop off are more valuable

@pawtrammell When AI makes outputs cheap the scarce parts become judgment taste and distribution. Labor share is not one bucket.

@pawtrammell Post singularity by its very nature is unimaginable, but effectively yes, we will never run out of things to do or solve.

@pawtrammell the labour-share argument also assumes demand stays legible. in rates terms, ai compresses old spreads, then invents entirely new duration risk for humans to warehouse.

@pawtrammell new bullshit jobs on the horizon

@pawtrammell The difference with AGI is that the production of any new categories of goods could be automatised.

@pawtrammell galaxy brain take. while we argue about post-singularity labour, tapbit is literally giving away world cup tickets right now 🎟️⚽ scan the QR in the pic, takes 30s. don't sleep on free vol

@pawtrammell When accounting is gone, I'll be at the opera

@pawtrammell new categories is the part nobody prices in. compute futures, agent labor, alignment-as-a-service didn't exist 3y ago. the wedge isn't human touch, it's the next abstraction layer.