AI Won't Cause Unemployment • Hacker News Board

Re: AI Won't Cause Unemployment

#231

Post by ShroudedNight » Sun, Mar 05, 2023, 9:48 AM UTC

twelve40 wrote: post #153

djs070 wrote: post #67

Earlier quoted context omitted.

His interpretation of “industries that are allowed to innovate vs those where it’s illegal” is laughable. Has Marc considered that there might be some other fundamental difference between the inflation-positive sectors of health, education, food and housing, vs the inflation-negative ones of TVs, cars and software subscriptions? I’ll give you a hint Marc: some of these are essential to life and society and some are n…

Because housing is essential to life does not mean it should double in price every ten years. Because education is valuable it doesn't mean the amount of student debt in the country needs to double like it has in the last two decades. At least in these two examples, these "industries" are flying off the charts in the last two decades not because of how essential they are, or how technologically progressive they are,…

> Because education is valuable it doesn't mean the amount of student debt in the country needs to double like it has in the last two decades.

If we change the basis of our model to financial pain, for things that are purchased with significant leverage, the cost going up as the interest rates go down would keep the financial pain relatively constant.

Re: AI Won't Cause Unemployment

#232

Post by TimPC » Sun, Mar 05, 2023, 9:49 AM UTC

pharmakom wrote: post #217

omnicognate wrote: post #191

Earlier quoted context omitted.

LLMs aren't new and there has been no sign of them being able to automate maths research so far. If they do, I will acknowledge them as AGI as it has been my goalpost for over 30 years. I've never moved it once, and I won't. Computers were invented in large part for this purpose. The hope was to have a machine that could spit out interesting theorems. It turned out that generating theorems was trivial (you can just e…

Interesting meaning interesting to humans? If an AI is doing all of the work then maybe it doesn't matter what we consider interesting? You gave a prompt asking for 2023 nostalgia pop music in the care home, and the AI found novel theorems to generate it for you, but it didn't even notice or care! Maybe there will be some hobbyist human Mathematians and the AI can throw them a clue once in a while, like a good murder…

Interesting generally means having some sort of sufficiently broad applicability to be useful and sufficiently narrow applicability to have enough explanatory power. If you are too narrow you don’t get applied in many situations and if you are too broad you’re generally too watered down to prove anything significant about the situations in which you apply. Some exceptions which makes everything harder.

1+1=2 is a theorem but a very uninteresting one because it applies exactly once.

Every number n has a double successor S(S(n)) is also uninteresting because while it applies more broadly it’s too watered down.

There are no numbers n with n greater than two such that x^n + y^n = z^n is extremely interesting.

Re: AI Won't Cause Unemployment

#233

Post by TimPC » Sun, Mar 05, 2023, 9:53 AM UTC

jonathankoren wrote: post #170

AnthonyMouse wrote: post #143

Earlier quoted context omitted.

> Just because there are new jobs doesn't mean the people whose jobs were automated out of existence will be able to take them. Yes it does, because those are the people whose labor will be available. They'll take whatever jobs they're qualified for, which may lower wages in those occupations until the lower price increases demand or reduces supply by driving anyone currently in those occupations who is qualified for…

> You get more production with less labor, so you on average need to do less labor to meet your cost of living. Citation Needed. This isn’t how it works out in practice though. If the predictions of the 1950s were to come true, we’d all be working 10 hours a week and bringing in the same wages. They weren’t wrong though. Efficiency did go up. It’s just that these efficiencies were never passed on to workers. Instead…

The predictions of the 1950s were wrong because people chose more stuff over less work. Partially because of the rising cost of land due to concentrated demand leading to needing “more home” in wage adjusted terms. Labour saw gains from increases in productivity and this didn’t really decouple until 1999 after which it mostly didn’t.

Re: AI Won't Cause Unemployment

#234

Post by ETH_start » Sun, Mar 05, 2023, 9:57 AM UTC

ChicagoDave wrote: post #212

This article is pure conservative BS. The stagnation of certain industries is because of tax policy and the insane greed injected into our healthcare systems when the HMO laws passed in the 1980's. I'm not even sure what pmarca is talking about here. It's like he's waving his arms, "look over here at my BS and don't look over there where AI is about to slam information workers." AI will absolutely change the way soft…

>>The stagnation of certain industries is because of tax policy and the insane greed injected into our healthcare systems when the HMO laws passed in the 1980's.

The growth of the healthcare bureaucracy is generally attributed to the increase in regulations:

https://www.athenahealth.com/knowledge-hub/practice-manageme...

While somewhat better designed to constrain costs, other countries' central economic planning in healthcare are also leading to runaway growth in healthcare spending as a percentage of GDP, and increasingly dire crises, e.g. Canada:

https://www.cbc.ca/news/health/doctor-shortage-cancer-video-...

What your comment illustrates though is how the dominant narrative will be shaped by the entrenched special interests who benefit from growing regulatory control, ensuring that Conservatism will remain the bogeyman even as the economy becomes increasingly centrally planned: https://ourworldindata.org/grapher/social-spending-oecd-long...

Re: AI Won't Cause Unemployment

#235

Post by TimPC » Sun, Mar 05, 2023, 9:57 AM UTC

jonathankoren wrote: post #193

AnthonyMouse wrote: post #178

Earlier quoted context omitted.

https://news.ycombinator.com/item?id=35027583 You're ascribing the evils of market concentration to automation. Automation isn't why housing and higher education are so much more expensive now than they were.

Yeah, I saw that post too, but didn’t want to pile on every one of your facile economic takes. You’re right, automation isn’t why housing and education are more expensive now. The increase of those is direct result of government policy choices encouraged by the voters. For housing, were taking zoning laws, and public objection laws. Higher education tuition increases for public universities is directly tied to cuts i…

A significant part of productivity gains went to workers up until 1999. There is no universal law that says productivity gains are inherently captured by shareholders. Of course if this is a possible option capitalists will choose it. I think this is a consequence of government policy where monopsony power prevents labour from getting a fair price due to only a handful of monopolists having the things that improve productivity.

Re: AI Won't Cause Unemployment

#236

Post by TimPC » Sun, Mar 05, 2023, 9:59 AM UTC

thatguy0900 wrote: post #136

AnthonyMouse wrote: post #128

Earlier quoted context omitted.

The reason technology doesn't net destroy jobs is that things that destroy jobs lower costs, which leaves people with more money in their pockets, which creates jobs. What net destroys jobs is things that unproductively increase costs. Resources that are wasted can't be used for productive labor that improves quality of life.

It seems to me all the money is just getting more and more concentrated. If Amazon succeeds in completely automating it's warehouses, who will see the bulk of that saved money, the customers or bezos?

Without a doubt the customers will see the bulk of the saved money. But Bezos capturing 5% of it will lead to an even more massive fortune. The minute any sizeable competitor follows suit it’s likely to be back to the typical roughly 2% retail margin with 98% of the savings going to customers.

Re: AI Won't Cause Unemployment

#237

Post by onion2k » Sun, Mar 05, 2023, 10:03 AM UTC

AnthonyMouse wrote: post #128

_448 wrote: post #101

This claim that AI won't cause unemployment baffles me. I understand the argument being made that new types of jobs will be created, but what the people making that argument miss is the massive level of up-skilling/re-skilling that is required of the population whose jobs are replaced. This has not worked out well. Just look at the communities in some poorer parts of the UK, who were relying on coal, mining and steam…

The reason technology doesn't net destroy jobs is that things that destroy jobs lower costs, which leaves people with more money in their pockets, which creates jobs. What net destroys jobs is things that unproductively increase costs. Resources that are wasted can't be used for productive labor that improves quality of life.

The reason technology doesn't net destroy jobs is that things that destroy jobs lower costs, which leaves people with more money in their pockets, which creates jobs.

This is true so long as two other things are also true;

- There needs to be something else for people to do instead. In the case of the industrial revolution power looms in factories replaced home weavers, but the factories were huge and could employ all the weavers and more. The factories created more jobs than they destroyed. In the case of AI that isn't immediately true, nor is it obvious that it's going to be true in the short term. So far we can only envisage the destruction of jobs.

- Secondly, the lowering of costs only puts money in people's pockets if the very wealthy don't hoard it for themselves. There are trillions sitting around doing nothing for the economy according to leaks like the Panama Papers. AI could potentially just make those numbers a bit bigger while everyone else starves.

In the long term I'm reasonably optimistic. I think society will organize around AI to let computers do things and we'll settle on an alternative model to 'time worked = money', whether that's something like UBI or artistic patronage or whatever really. I think that's 50 years away. We'll only get to that because in the short term it's going to be painful though.

Re: AI Won't Cause Unemployment

#238

Post by dogcomplex » Sun, Mar 05, 2023, 10:06 AM UTC

LeroyRaz wrote: post #16

I think there is real difference between a) devaluing someone's work, and b) and voiding work of value. What makes AI different from previous technological innovations is that it seems to have the potential to make peoples labor worthless, rather than merely devalued. To give examples, if your writing is worse than ChatGPTs then your writing is worthless. If your art is worse than stable diffusion then your art is wo…

> I do not think our current social policies are ready for that.

Understatement. (And also, they weren't really ready for the previous waves either)

The upside is we exist as both producers and consumers - and while our value is about to go down (or be voided entirely), so is the price of all the things we want/need. Can't afford a house because the Atlas robot outperformed you at construction? Well, get a few buddies to pool funds to buy an Atlas robot yourselves and get it to build your houses. The barrier of entry is similarly dropping fast. It's gonna be a weird, sad and awkward short-to-medium term as the fluctuation of which drops first - employment or consumer prices - plays out, especially if government doesn't step in to protect people. But long-term I think it's hard to argue that there's not going to be an enormous increase in overall quality of life. (barring some sort of annihilation event...)

Re: AI Won't Cause Unemployment

#239

Post by hef19898 » Sun, Mar 05, 2023, 10:07 AM UTC

_448 wrote: post #101

This claim that AI won't cause unemployment baffles me. I understand the argument being made that new types of jobs will be created, but what the people making that argument miss is the massive level of up-skilling/re-skilling that is required of the population whose jobs are replaced. This has not worked out well. Just look at the communities in some poorer parts of the UK, who were relying on coal, mining and steam…

If we take the industrial revolution as an example, yes, it is true that on average, at the minimum and as a whole indistrialized nations and societies are better off after it then before.

Also true, the period between before and after sucked, killed and ruined the health of generations, affected mostly those whose skills weren't as valuable as before anymore and the poor. Ignoring this change, and the social impact it has, is heartless capitalism at its worse. And during the industrial revolution, it was only addressed due to labor organisation and action, unions and public pressure so high that governmebts couldn't ignore it anymore. And it gave birth to communism.

Not sure if AI amounts to the same level of impact as industrialisation did, but f it does maybe it it would be sound policy to address the resulting social problems before they are to severe to ignore. And no, I am not talking about six figure salaries tech "workers", I talk about all those under paid and close to minimum wage blue collar real workers.

Re: AI Won't Cause Unemployment

#240

Post by intellectronica » Sun, Mar 05, 2023, 10:19 AM UTC

This argument can be backtested, sort of. AI is novel technology, but information technology in general has been on a wild run for now 2~3 decades. One would expect the advances in information technology to result in significant productivity increases that would affect the entire economy and likely result in unemployment. None of that has happened. Productivity increases from digital technology growth are really hard to detect. In fact the same period that featured the most significant growth in digital technology is also the period where productivity increases in developed economies stagnated.

Pmarca's explanation predicts that - if large and growing parts of the economy are blocked to technological improvements then no matter how much technology improvements we develop it won't affect the economy in the aggregate. This was true for the last three decades. Hard to see why it wouldn't be true also going forward.