Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Generative AI is not replacing jobs or hurting wages at all, say economists

theregister.com

Post reply on HN
Print view View original

371–380 of 503 posts

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#371

Post by jhp123 »

bhelkey wrote: post #347
jhp123 wrote: post #313

The thing about AI is that it doesn't work, you can't build on top of it, and it won't get better. It doesn't work: even for the tiny slice of human work that is so well defined and easily assessed that it is sent out to freelancers on sites like Fiverr, AI mostly can't do it. We've had years to try this now, the lack of any compelling AI work is proof that it can't be done with current technology. You can't build on…

> [LLMs] won't get any better. Can you pinpoint the date which LLMs stagnated? More broadly, it appears to me that LLMs have improved up to and including this year. If you consider LLMs to not have improved in the last year, I can see your point. However, then one must consider ChatGPT 4.5, Claude 3.5, Deepseek, and Gemini 2.5 to not be improvements.

Sept 2024 was when OpenAI announced its new model - not an LLM but a "chain of thought" model built on LLMs. This represented a turn away from the "scale is all you need to reach AGI" idea by its top proponent.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#372

Post by jononor »

jrflowers wrote: post #356
jaredklewis wrote: post #334

Earlier quoted context omitted.

The headline is about wages and jobs. It’s very possible that AI could result in time savings of 50% of work hours in a week and still have no impact on wages or jobs. This is because the economy is not a static thing. If one variable changes (productivity), it’s not a given that GDP will remain constant and jobs/wages will consequently be reduced. More likely is that all of the variables are always in flux, reacting…

> It’s very possible that AI could save 50% of work hours in a week and still have no impact on wages or jobs. I like this sentence because it is grammatically and syntactically valid but has the same relationship to reality as say, the muttering of an incantation or spell has, in that it seeks to make the words come true by speaking them. Aside from simply hoping that, if somebody says it it could be true, “If every…

Doubling the wages is not going to happen... But it could be that output gets doubled, at the same personell cost (jobs*wages). Ref Jervons Paradox.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#374

Post by datpuz »

bredren wrote: post #327
1vuio0pswjnm7 wrote: post #240

Earlier quoted context omitted.

"I'm generally an AI skeptic, but it seems awfully early to make this call." What call. Maybe some readers miss the (perhaps subtle) difference between "Generative AI is not ..." and "Generative Ai is not going to ..." Then first can be based on fact, e.g., what has happened so far. The second is based on pure speculation. No one knows what will happen in the future. HN is continually being flooded with speculation,…

> According to the study, "users report average time savings of just 2.8 percent of work hours" from using AI tools. That's a bit more than one hour per 40 hour work week. Could be data is lagging as sibling comment said but this seems wildly difficult to report on a number like this. It also doesn't take into account the benefits to colleagues of active users of LLMs (second order savings). My use of LLMs often mean…

IMO, if you're gaining a significant amount of productivity from LLMs in a technical field, it's because you were either very junior and lacked much of the basic knowledge required of your role, or you performed like you were.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#376

Post by freejazz »

mac-attack wrote: post #298
1vuio0pswjnm7 wrote: post #240

Earlier quoted context omitted.

"I'm generally an AI skeptic, but it seems awfully early to make this call." What call. Maybe some readers miss the (perhaps subtle) difference between "Generative AI is not ..." and "Generative Ai is not going to ..." Then first can be based on fact, e.g., what has happened so far. The second is based on pure speculation. No one knows what will happen in the future. HN is continually being flooded with speculation,…

> In contrast, this article, i.e., the paper it discusses, is is based on what has happened so far. What happened in 2023 and 2024 actually Nitpicky but it's worth noting that last year's AI capabilities are not the April 2025 AI capabilities and definitely won't be the December 2025 capabilities. It's using deprecated/replaced technology to make a statement, that is not forward projecting. I'm struggling to see the…

Well, it's a given that the sun is shining when it is out. Not so much with the AI.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#377

Post by cornholio »

ringeryless wrote: post #64
cornholio wrote: post #31

Earlier quoted context omitted.

Probably the first significant hit are going to be drivers, delivery men, truckers etc. a demographic of 5 million jobs in US and double that in EU, with ripple effects costing other millions of jobs in industries such as roadside diners and hotels. The general tone of this study seems to be "It's 1995, and this thing called the Internet has not made TV obsolete"; same for the Acemoglu piece linked elsewhere in the.…

LLMs are the least deterministic means you could possibly ever have for automation. What you are truly seeking is high level specifications for automation systems, which is a flawed concept to the degree that the particulars of a system may require knowledgeable decisions made on a lower level. However, CAD/CAM, and infrastructure as code are true amplifiers of human power. LLMs destroy the notion of direct coupling…

If you give the same subject to two different journalists, or even the same one under different "temperature" settings, say, he had lunch or not, or he's in different moods, the outputs and approaches to the subject will be completely different, totally nondeterministic.

But that doesn't mean the article they wrote in each of those scenarios in not useful and economically valuable enough for them to maintain a job.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#378

Post by asadotzler »

sharemywin wrote: post #193
atrus wrote: post #162

Earlier quoted context omitted.

You're falling into the mistake of "correct" or "lied" though. Being wrong isn't lying .

which is interesting because AI doesn't have intent and there is incapable of lying.

Of course it has intent. It was literally designed to never say "I don't know" and to instead give what ever string of words best fits the patter. That's intent. It was designed with the intent to deceive rather than to offer any confidence levels or caveats. That's lying.

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#379

Post by notahacker »

decimalenough wrote: post #88

I'm generally an AI skeptic, but it seems awfully early to make this call. Aside from the obvious frontline support, artist, junior coder etc, a whole bunch of white collar "pay me for advice on X" jobs (dietician, financial advice, tax agent, etc), where the advice follows set patterns only mildly tailored for the recipient, seem to be severely at risk. Example: I recently used Gemini for some tax advice that would…

A friend of mine is working on a "RAG for tax advisers" startup. He's selling it to the tax accountants, as a way for them to spot things they wouldn't otherwise have the time to review and generate more specialist tax advisory work. There's a lot more work they could do if only it's affordable to the businesses they advise (none of whom could do their own taxes even if they wanted to!)

Jevons law in action: some pieces of work get lost, but lower cost of doing work generates more demand overall...

Re: Generative AI is not replacing jobs or hurting wages at all, say economists

#380

Post by asadotzler »

Workaccount2 wrote: post #207
ozgrakkurt wrote: post #105

Earlier quoted context omitted.

It can’t replace a human for support, it is not even close to replacing a junior developer. It can’t replace any advice job because it lies instead of erroring. As an example if you want diet advice, it can lie to you very convincingly so there is no point in getting advice from it. Main value you get from a programmer is they understand what they are doing and they can take the responsibility of what they are develo…

I trust cutting edge models now far more than the ones from a few years ago. People talk a lot of about false info and hallucinations, which the models do in fact do, but the examples of this have become more and more far flung for SOTA models. It seems that now in order to elicit bad information, you pretty much have to write out a carefully crafted trick question or ask about a topic so on the fringes of knowledge…

I get wrong answers for basic things like how to fill out a government form or the relationship between two distant historical figures, things I'm actually working on directly and not some "trick" to get the machine to screw up. They get a lot right a lot of the time, but they're inherently untrustworthy because they sometimes get things subtly or catastrophically wrong and without some kind of consistent confidence scoring, there's no way to tell the difference without further research, and almost necessarily on some other tool because LLMs like to hold onto their lies and it's very difficult to convince them to discard a hallucination.
Post reply on HN