Comment by mbgerring
2 days ago
Can someone in this thread who says “the kids must be wrong” give an actual optimistic case for AI? Because as far as I understand it, the “optimist” case for AI is that LLMs become God and wipe out human life as we know it entirely, and replace it with a transcendent post-human intelligence. And in the meantime, we’ll have a permanent underclass that will be kept alive on some kind of subsistence UBI. That seems to be the “good” outcome that e.g. OpenAI is playing for. I don’t understand why any of you think that’s good or positive or desirable.
> that will be kept alive on some kind of subsistence UBI
Setting aside whether a permanent underclass will be an outcome or not - is it not a bit incompatible to simultaneously believe that all jobs will be gone and that a subsistence UBI is necessarily very bad?
The way I see it, if strong AGI really replaces all jobs, then even a subsistence level UBI (by the new post-AGI standards) will be a world with ubiquitous resources and post-scarcity wealth where people pretty much spend their time how they want. Yes, perhaps it might be an "underclass" in the sense that Musk and Altman meanwhile settled Mars with some privatized space colonization, but I might still be orders of magnitudes richer than I am today - so why should I care, except for status games?
It's basically like someone in the pre-industrial age complaining that the industrial revolution will lead to the permanent underclass of people who currently live on welfare in developed countries. But this is surely a wonderful outcome from the pre-industrial peasant's perspective - even poor people in the western world live better than middle age kings!
> post-scarcity wealth where people pretty much spend their time how they want
In a true post-scarcity society where Musk et al are off colonizing Mars while you're stuck on Earth on UBI, doing what exactly? In such a future, AI has automated boring chores but also everything else. Art, movies, cooking, everything you might find enjoyable. So lots of free time to do what? Work on yourself? Nobody will care. Engage with your hobbies? Nobody will care to see them, so unless you're the kind of person who enjoys their inner life without interaction with others, be aware nobody will read your AI novel nor watch your AI movie, because they can make one specifically tailored for themselves.
To me this seems like a hellish future, a mix of "Farenheit 451" with people staring at wall screens (only AI-tailored for them) or the "basic income but restless" people from Earth in "The Expanse". It's even worse than "A Brave New World" because human workers will be mostly unneeded. Their basic necessities covered, but nothing for them to do, no real struggle other than boredom. Any challenges that remain must be artificially self-imposed, because the real challenges will be for a chosen few.
> Work on yourself? Nobody will care. Engage with your hobbies? Nobody will care to see them
But this is already the case now, no? The only person you do self-improvement or hobbies for is yourself or your closest social circle. We are already worse at everything than the professionals in the respective fields and also do not despair about it.
> To me this seems like a hellish future, a mix of "Farenheit 451" with people staring at wall screens (only AI-tailored for them) or the "basic income but restless" people from Earth in "The Expanse".
I think a more optimistic take people usually bring up is the Culture series. But yes, of course ultimately it all boils down to us being obsolete. That does not give me worse existential angst than life (to which there's no real point) already does, however, so I don't see it as worse than the status quo.
7 replies →
> It's basically like someone in the pre-industrial age complaining that the industrial revolution will lead to the permanent underclass of people who currently live on welfare in developed countries. But this is surely a wonderful outcome from the pre-industrial peasant's perspective - even poor people in the western world live better than middle age kings!
Wow, it took more than a century to get to live better than middle age kings. Do you think the people that lost their means to make a good living in 1800s care that we live so much better today ? It is incredible how so many tech people lack empathy for how regular people think or want to live.
I liked the Culture series too, but how they got to the presented post scarcity world is never described. How many generation lived a worse life than their predecessors? Do you think the current or future bi and trillionaires are willing to pay everyone a decent wage to live through this transition period ?
> Wow, it took more than a century to get to live better than middle age kings. Do you think the people that lost their means to make a good living in 1800s care that we live so much better today ? It is incredible how so many tech people lack empathy for how regular people think or want to live.
I noticed that comes more from a place of not imagining themselves being the subject of the possible terrible consequences.
Many tech people think in abstract terms, they look back in history without thinking much about how the life of a normal person during a major transitional period was impacted, it's just a sequence of facts, not a collection of human stories.
It's sad because it completely detaches many of these folks from having empathy, yes, change is the only constant but if our aim is to progress as a species we should also be progressing on how to make inevitable changes less miserable for those impacted. I see a lot in tech people the thinking of technological advancement for the sake of technological advancement, not for building a better world for every human, humans tend to get in the way of major technological changes so in their minds they prioritise the advancements without caring much about the human aspect.
It's quite baffling to me because those are usually smart people, I'd expect smart people to have better holistic thinking.
In my view, UBI basically puts me, an upstanding citizen with hard skills that AI made obsolete, on the same playing field as the average junkie off the street in SF. Why on earth would I want that? People are different, the modern economy is a great stratification mechanism at putting you near people of similar conscientiousness, and getting rid of that is a recipe for misery.
>Can someone in this thread who says “the kids must be wrong” give an actual optimistic case for AI?
The optimistic case is technological deflation. Where goods and services become so cheap, you don't need a lot of money to afford them. If you can have a robot sort packages like,
https://tech.yahoo.com/ai/articles/human-intern-beats-figure...
Why have a human do that? I don't think there's a person alive whose life goal is to sort packages. A human will lose a job, but only a job they accepted because the human needed money. Well if the package sorting drops the price of things, they don't need as much money. Now if every job is robotic, everything becomes cheaper to the point we don't need money for very many things at all.
That's the optimistic case.
OK, and in a world where this technology is broadly available and not controlled by 4-5 companies with an unassailable capital moat, I can see how this could be a good outcome. But that’s not the situation we’re in.
For me to understand this as an “optimistic” case, I need to understand why people believe that absent a need for human workers, there will be any incentive for the people who control all of the capital to keep people alive.
>controlled by 4-5 companies with an unassailable capital moat
This has been the situation for CPUs for decades. We now carry a 1980 cray super computer in our pockets.
>For me to understand this as an “optimistic” case, I need to understand why people believe that absent a need for human workers, there will be any incentive for the people who control all of the capital to keep people alive.
They won't need to engineer a Terminator style genocide of mankind. Man will kill each other in another war, like we always do. Biological imperatives mean we all kill each other until resources are abundant for those left standing. Then the winners of that war have a baby boom, their children are boomers for the next 80 years, and we start all over again.
The optimistic case says the robots are so freakin' good, they create the abundance for us without the need for the killing.
1 reply →
The pessimistic case is based on the misconception that AI is some kind of a superhuman. Our current AI models are trained on human data, which has an unfortunate side effect which causes them to think and behave like a human. But as soon as we learn to train them without human data, we find out that AI is just a supercalculator, and it won't have any own will or agency.
Will and agency are primal biological instincts, which a pure intelligence doesn't have. It doesn't want or need anything. Therefore it won't act.
A superintelligence with human primal instincts would be scary indeed, but obviously we don't want to build that.
What you mention of training without human data seems to me an impossibility. Unless you're talking about going back to programming an AI via traditional methods rather than relying on machine learning (which might not be impossible, hard to prove it as such at least).
I don't think you can divorce intelligence from all biological aspects and just get computational power. It's an interesting question though..
Increasingly AI's are trained using reinforcement learning [1] so these are not really human tasks but things like trying to prove theorems, play games, solve code problems and getting feedback from compilers and similar. A lot of the early pop science coverage of AI was around the ideas of "data walls" and constrains of human data, most of which just wasn't really true or long term true anyway.
[1]. https://www.dwarkesh.com/p/dario-amodei-2
Well.....there is the 'lump of labor' fallacy that states that there is a fixed amount of work to be done in the economy. There is also the thought that a super intelligent AI would go the route of the Buddha having run all simulations under the virtual Bodhi tree and become benevolent. But most people think its going to be the Terminator so.......
Did you ever watch Star Trek: The Next Generation? The current trajectory is like the Ship's Computer. It know everything humanity has learned and can do a lot. But it can't explore and lacks desires and agency. That's why they made a big deal about the character Data being an entirely new kind of AI. Of course Star Trek has a very different economic system and there is a book called Trekenomics about that. So optimistically people live for themselves and don't persue labor they despise. Half of Americans hate their jobs and live for the dream of retirement when they get to actually do what they want.... But they don't have the same energy anymore.
Nonsense. Nobody's gonna bother with the subsistence UBI.
This is what bugs people. We can tell the part they're bullshitting about is the promise of a subsistence UBI. No wonder people boo.
I’m know it’s better in some other countries, but in the U.S., we can’t even agree that all people with jobs should have health insurance.
That’s even a bit optimistic. We can’t agree that all people with full time jobs should be able to afford the basics for their own survival.
1 reply →
Elon Musk is busy arguing to massively cut social security because it's fiscally unsustainable. He's also claimed that AI will create so much wealth that 'everyone can have a penthouse if they want'. These beliefs do not seem consistent, but the instinct to fight taxation is extremely consistent.
health - issue detection, cancer treatments, new drugs creation, etc..
To everyone replying here: if you think that the owners of the machines will have any incentive to keep you alive once they no longer need you, or your reading of history is that kings and capitalists have ever willingly made concessions to their power out of benevolence, you are going to have a bad time.
Sure. The machine gods are benevolent gods who care deeply for their creator-species. We are freed from labour and troubles into a paradise, to eat peaches and cream and make love under the sun. Rich or poor, we'll all be emperors of our domain, free to do as we please. Our lives keep getting better and better with technological progress, at least in the scope of our social-capitalist system. They will only get better until they end.
I need you to understand that if you actually believe this, normal people think you are an evil lunatic.
I don't think you need to worry yourself so much on my account friend. You asked, I told. Let's keep it at that. Also feel free to keep the name-calling to yourself.
You asked for an optimistic case and he gave you one. One thing I really like about LLMs is that they don't engage in this type of petty deceit where they ask a question and then insult you for answering.
5 replies →
I like this version. Never understood the assumption that it all had to go the route of Terminator.
yes, all hail the Morlocks as we Eloi live in peace.
HG Wells really did have a time machine!
Yep, the Time Machine minus the fear, the cannibalism, the suffering, the apathy, the reduced capacity for engagement etc are essentially the best-case scenario, which takes us out of HG Wells enough for it not to matter much as a cautionary tale.
We know how all the plumbing works so at least we'll get eaten last?
I'm more "the kids are showing healthy skepticism of corporate dystopia but AI is vital" camp, here is my argument:
1. Yes the risk of AI corporate/authoritarian dystopia is HUGE, we'll have to fight for our rights MANY times this century. Transcendental AI takeover is probably less of a risk than humans in power using armies of robots and Stasi-AI surveillance.
2. Our current economy is bs and the last century of 'relative prosperity' was a bit of luck + tech and population boom + globalist exploitation and massaging debt. We've tried variations of capitalism, socialism, communism, there doesn't seem to be a silver bullet.
3. AI is not like other tech, and tech does not 'create jobs', it creates business opportunities which up until now have always translated to jobs. We've never had a "drop-in replacement" for a human employee, it could replace anything from 40 to 99% of jobs.
Those are the risks, the potential rewards are:
- OpEx converted to CapEx making almost any kind of business extremely efficient
- Nobody has to spend weeks away from family or risking their health in dangerous or degrading jobs
- Extremely cheap housing and infrastructure with everything from mining to construction to maintenance automated. Fixing the broken window effect of rundown neighbourhoods and generally increasing quality of life
- Almost nobody needs to commute, or do all the other things around commuting, vastly reducing transport, congestion and pollution
- Food can be grown in better ways, even at home, with less mono-cropping, pesticides and waste. Your robot can weed by hand, work the land 24/7 and with the combined experience of millions of farmers, botanists etc
- Healthier society, no need for convenience food if your robot can cook and clean, and it can make far tastier traditional food than McDonalds
- Many products can be made at home or locally. Mass production favoured big dumb machines but a robot can build you a table exactly how you want it, with appropriate materials rather than commoditising everything down to shitty MDF off-gassing formaldyhide. You don't have time to pick through recycled wood - your robot does
- Our existing road network can have far higher capacity because barely anyone needs to commute and idiots don't hold up traffic or drive distracted. Streets aren't jammed with parked cars, taxis instantly have 20% extra capacity as they don't need to carry a driver. We may even get rid of or severely reduce traffic lights, not to mention safety
- Anything in your life that involves expensive repairs or buying more dumb shit is improved, every robot is a plumber with 100 million job experience, so many problems are solved with a machine that combines cheap labour and wide expertise
"Oh but humans need purpose" I just don't think 90% of jobs provide purpose. Purpose is raising kids, spending time with friends and family, working on some project, art, community improvement - it's absolutely insane we spend so much time working on bs.
Even just one of these things coming true is revolutionary - we have turned into fat commuter drones stress eating stuck in traffic thinking about some abstract spreadsheet report so far removed from reality but stealing our sleep and peace. AI isn't the problem here its corporate greed and concentration of power that AI could give
I feel like the main issue here is that it does seem like in the current trajectory the job loss outcome is going to happen before any of these potential really good outcomes. I'm down for a utopian future, but I don't have to want to spend 10 to 20 years in a depressing unemployed hellscape before I get there.
Agreed, there is one simple solution that always works albeit limited and temporary. Jefferson said it, Luigi did it, many are thinking it - but we must ensure it does not turn into a mass purge, just limit to maybe a few parasitical PE players and everyone else will fall in line for at least a few years.