The Education of a Doomer

ArcHound

The issue with devaluation of human work is that the automation doesn't have to be real. All you have to do is make nearly all employers agree that a particular job can pay less. And that's the case today - new AI being a sufficient programmer replacement is enough of a statement to do layoffs with record profits without impact on stock prices.

rahulyc

I'm not sure what happens (I think no one does), but with every other technology, I can opt-out if I don't want to use it. Even if I get unemployed, I don't see how I can opt-out when it comes to everyone everywhere using it. Governments historically have considered humans as assets - it's even called "human capital", and they require us for tax revenue and to run the state. What happens when we aren't needed? Even if the AI is magically aligned to the state, I don't see how I end up the winning side of that. It's not like we're getting a voice in how this is shoved down our throats today. Why would it get any better when AGI can do everything humans can do? What still gives us a voice in the world?

paulpauper

I’ve never had this worry. When I was funemployed last year, I spent my time reading books and writing code and hanging out with friends. If the future is an infinite UBI-funded vacation, I know what I’ll do. “Before the Singularity, read books and throw house parties; after the singularity, etc.” yeah it's obvious the author has a really high IQ who can keep himself busy with intellectual pursuits and a nice social life of other smart people . His experience will not map to the general population, who find themselves with idle time. Also, a barebones UBI will be hardly sufficient to cover other expenses unless the AI also solves housing, healthcare, insurance, childcare, transportation etc.

ankithg

One thing the doom debate tends to skip: the failures that hurt people right now are boring, not exotic. I spend my time probing AI systems that are already deployed and talking to real users, and the actual failure modes are mundane - a system that reassures when it should escalate, follows its script while the situation changes, sounds confident the whole way through. No superintelligence required, just an ordinary model given authority and no tripwires. If the conversation centered on the systems already in production instead of hypothetical future ones, a lot of the disagreement about timelines would matter less. The dangerous stuff is already here; it's just unglamorous.

throwawayk7h

> if humans are economically useless, the state does not need them. Why pay out UBI to people who have neither economic nor political power? Forgive me if I'm naïve. But wouldn't people just vote for UBI in this case?

slopinthebag

Actual AGI might be really bad but thankfully we are not even close to building it and I doubt I’ll see it in my lifetime (imo ofc)

perching_aix

TL;DR: the same tired points you've read a million times already "It might be slop but at least it's human slop!" - would the author argue in turn, probably: > People used to argue online, flame each other, get angry. And that wasn’t good, but at least it was human. [0] Another perspective though could be that one's thoughts were never meant to be guaranteed high-entropy curio for others to encounter, so they're really under no obligation to be found "original", if that even means anything. [1] And sure, that is true. Just also not any more true than the former. [0] emphasis theirs, I don't even need to mock it like this; the secondhand embarrassment comes built-in [1] this is to say, these may be genuinely candid and original thoughts from the author, that simply happen to coincide with the many thoughts already expressed in this genre of writing

ghostly_s

> You might object: if we’re all jobless, who’s paying for everything? This is trivially answered: the state acts like the heart, taxes are venous blood and welfare is oxygenated arterial blood. The government pays Raytheon for missiles, the money cascades down the economy through factories, aluminium smelters, mines, transport companies, all staffed by AIs buying and selling from each other. The government takes a cut of all economic activity, pays out welfare, the unemployed masses buy food and pay rent, the supermarkets, farms, logistics network, etc. are all staffed by AI. I don't think this person has a firm enough grasp of economics to be writing on the subject.

skybrian

> What’s the last major life decision you made without at least consulting an AI? This seems like asking what's the last major decision you made without at least doing a web search. There's nothing wrong with doing a search. The trouble is when you rely too much on what you read, without checking it.

aranelsurion

> what if GPT-3 hadn’t worked? What would we have to look forward to, in this alternate 2026? One would hope maybe boring things like healthcare, better living conditions, better functioning societies, and so on. Not as exciting as flying cars and thinking machines for sure.

underlipton

>Before LLMs, what was there to be excited about in the near future? Solar power? Self-driving cars? The energy transition? It’s laughable. I, for one, would have welcomed our green energy revolution overlords. Also, VR/AR, before the Great Powers deep-sixed any real efforts. Growing up on .hack, GitS, and SAO, it really is a bummer. How come I must know, where the techbros lead me to go?

akoboldfrying

All it would actually take to avert a jobs crisis is for the per-token price of LLM generation to increase to the point where humans are competitive in dollar terms for intellectual work. This could happen all by itself if the major LLM players coordinated (i.e., formed a cartel), or if they were forced to to recoup investments (we all know this is coming to some degree , at least for the western companies). Alternatively it could be forced into place if ~all major nations' governments start taxing AI use (which seems eminently a good idea to me in any case). There's still the issue of local LLMs, but this could be addressed with taxes on GPU hardware, etc.

erelong

> Automation has been good I think the only main critique that comes to mind is the primitivists, who argue that automation has massively destroyed our freedom and psychologically destabilized us (among other things); you really can't meaningfully opt out of whatever technology exists and it forces someone to use it somewhere. It also ultimately takes away certain freedoms that previously existed and creates lots of new problems that didn't exist before (like increased mental health problems in industrialized nations). And so this problem has already happened with the Industrial Revolution - AI is basically just a continuation of these problems, not perhaps different in kind but degree. > Alignment is an issue to me, but any currently existing technology can also be misused to some extent > Control again we're not in control of current technology, new things are developed that other people have no control over but just "adapt" to > by letting the AI write, you are giving up most of what makes up thinking I guess I haven't been viewing things this way but have been filtering ways I anticipate AI should be used and not used as. In other words, I think it's a great use of AI to share a prompt and AI output as a way of dialogue "writing". For example, I might want to give you 10 ideas on something to bake, and AI outputs 10 ideas that are actually valuable, so sharing the list does add value. But in how I would "write" with AI, it would be clear where I am prompting and where the AI takes over. I also think that a person should read the AI output and essentially indicate points of agreement or disagreement especially if there is some serious defect in the AI output. > The discourse around software engineering has gotten dumber Good and bad; it's enabled people with less SE experience (or intelligence even) to talk more about SE and become a meaningful participant in the process of software creation. But it takes less specialized skill for some of these things than some of the previously developed skills, and that can have a different feel to it. > Deferring to AI Again I guess I am filtering a bad use of this. AI frequently gives "good" answers that I would have manually hunted for previously on a search engine, and I still can sift through what it outputs. So I think it's great as a brainstorm tool that I somewhat freely choose from. I can understand how other people may not be using it in the same way as I use it though, more like an aide rather than a smarter entity I am taking orders from. > Stagnation I guess it's ironic because although I also focus on "stagnation", I guess there is indeed a lot of actual "progress" going on. It's more of a political issue perhaps than a technological issue.