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

AI used well can restore middle class jobs [pdf]

nber.org

Post reply on HN
Print view View original

131–140 of 143 posts

Re: AI used well can restore middle class jobs [pdf]

#131

Post by bumby »

pydry wrote: post #130
bumby wrote: post #108

Earlier quoted context omitted.

I would agree, but your statement makes it sound as if they are disjointed. You need to have both. (I would also push back a bit that a PhD adds immense value. Many that do get paid well via IP and start-ups, but that’s a better interface with the economy. There are many in my experience that add little more than a credential)

>You need to have both. For employment the most it does is set a ceiling on wages. Pushing up your ceiling doesn't necessarily push up your wages. >I would also push back a bit that a PhD adds immense value. Of course, that's a common opinion. Blue sky research is vastly underrated in the west. This is partly why the Chinese are overtaking us.

>Blue sky research

I’m not discounting fundamental research in the way you may be implying. I’m just saying it has little economic value until it has an application. By definition, it doesn't have an economically marketable use, so it can't have economic value.

I also don’t think China is overtaking the US on a research front. They publish a lot (although I would question how much 'blue sky' vs. derivative), but that’s a bad metric for a lot of reasons.

Re: AI used well can restore middle class jobs [pdf]

#132

Post by ben_w »

smallmancontrov wrote: post #107
ben_w wrote: post #81

Earlier quoted context omitted.

> There are better definitions. I'm a linguistic descriptivist, not a prescriptivist. This is how I think most use the term, even if there's a more useful term of art in economics or in politics. > You said Voldemort's name! Now he will surely haunt us One might even say that a spectre is haunting Europe… ;P > So the composition of incentives switches from worker-like to owner-like somewhere around $10MM net worth, a…

Sure, and I see the wisdom in it: everyone gains 50 IQ points when they place a bet, so private ownership of the economy means that every section of the economy is owned by a responsible skin-in-the-game party incentivized to maximize value. A few simple rules create distributed intelligence. Cool! But I also see the dark side. The economic notion of "value" being maximized is actually "wealth weighted value," a tiny…

> But I also see the dark side.

Indeed; if there were no dark side, Marx would have had nothing to write about.

The difference between "money" and "value" can also be found in the difference between what is measurable and what is important is noted in Goodhart's law, which was originally specifically about two different measures of money and why putting pressure on them specifically for economic reasons was not having the anticipated result: https://en.wikipedia.org/wiki/Goodhart%27s_law

I hope that the successor to Smith and Marx will account both for Goodhart and for Nash equilibria/game theory, both of which were developed and formalised over a century after the Communist Manifesto and around two centuries after The Wealth of Nations.

Re: AI used well can restore middle class jobs [pdf]

#133

Post by pydry »

bumby wrote: post #131
pydry wrote: post #130

Earlier quoted context omitted.

>You need to have both. For employment the most it does is set a ceiling on wages. Pushing up your ceiling doesn't necessarily push up your wages. >I would also push back a bit that a PhD adds immense value. Of course, that's a common opinion. Blue sky research is vastly underrated in the west. This is partly why the Chinese are overtaking us.

> Blue sky research I’m not discounting fundamental research in the way you may be implying. I’m just saying it has little economic value until it has an application. By definition, it doesn't have an economically marketable use, so it can't have economic value. I also don’t think China is overtaking the US on a research front. They publish a lot (although I would question how much 'blue sky' vs. derivative), but tha…

>it has little economic value until it has an application

That's a bit like saying that employees create very little value until a product is sold. Realized value != created value.

It is precisely the lack of competition on this front (research) for decades and the abstract nature of the value created that has led it to be so badly undervalued.

Re: AI used well can restore middle class jobs [pdf]

#134

Post by bumby »

pydry wrote: post #133
bumby wrote: post #131

Earlier quoted context omitted.

> Blue sky research I’m not discounting fundamental research in the way you may be implying. I’m just saying it has little economic value until it has an application. By definition, it doesn't have an economically marketable use, so it can't have economic value. I also don’t think China is overtaking the US on a research front. They publish a lot (although I would question how much 'blue sky' vs. derivative), but tha…

>it has little economic value until it has an application That's a bit like saying that employees create very little value until a product is sold. Realized value != created value. It is precisely the lack of competition on this front (research) for decades and the abstract nature of the value created that has led it to be so badly undervalued.

That's exactly the point. To put it in more HN terms, the best-engineered software has no economic value until it is put into use/sold. Created value != economic value. So what is the economic value if its never realized? Nothing. The entire distinction is that there is a difference between types of value; my post focuses on economic value because that is what's related to someone's pay.

If you could quantify fundamental research's value to the economy (without hindsight), it ceases to be fundamental research. Again, by definition, fundamental research does not have application. If it has no application, it has no interface to the economy. I'm not saying R&D or fundamental research has no potential economic value. It has no current avenue for realized economic value. And our system is incentivized for short-term gain. You could be a magnificent fundamental researcher, but if your work doesn't add to the economy in a relatively short timeframe, you probably won't be paid very well compared to a lesser researcher who does.

Re: AI used well can restore middle class jobs [pdf]

#135

Post by bumby »

golergka wrote: post #114
pydry wrote: post #96

Earlier quoted context omitted.

If they have any interest in the country, they have taxable activity. Whether the government wants to tax it is a different question. Usually they do not.

That's how you make people lose interest in the country — exactly what OP is complaining about. If you want to stop brain and capital drain, taxation should be the last thing on your mind.

There's a tipping point though, right? I think the right question is, how do we find the correct threshold?

Otherwise, there will be an asymmetry in which the "brains and capital" will get all the benefits of taxes with little to no skin in the game. That implies society serves the economy and not the other way around.

Re: AI used well can restore middle class jobs [pdf]

#136

Post by pydry »

bumby wrote: post #134
pydry wrote: post #133

Earlier quoted context omitted.

>it has little economic value until it has an application That's a bit like saying that employees create very little value until a product is sold. Realized value != created value. It is precisely the lack of competition on this front (research) for decades and the abstract nature of the value created that has led it to be so badly undervalued.

That's exactly the point. To put it in more HN terms, the best-engineered software has no economic value until it is put into use/sold. Created value != economic value. So what is the economic value if its never realized? Nothing. The entire distinction is that there is a difference between types of value; my post focuses on economic value because that is what's related to someone's pay. If you could quantify fundame…

>software has no economic value until it is put into use

This is exactly my point. This ^ is categorically false.

>If you could quantify fundamental research's value to the economy

It seems like you're committing the https://en.wikipedia.org/wiki/McNamara_fallacy

Re: AI used well can restore middle class jobs [pdf]

#137

Post by Eddy_Viscosity2 »

lancesells wrote: post #110
Eddy_Viscosity2 wrote: post #88

Earlier quoted context omitted.

If every single high school graduate then got a bachelors degree, would their median earnings match the median earnings of the of the bachelors degree group now? It doesn't seem likely. Otherwise we can just mail every single person a bachelors degree certificate in the mail and watch the salaries increase, productivity sore, the economy on permanent boom!

Don't you think it's correlated to the person having an education and skillset that someone who didn't earn the degree would normally not have?

Of course I think that education matters and is the important thing here. But employers require a degree (i.e. the piece of paper) so that's the thing they are using as a proxy and so that's the thing that may decide whether you get a job (or even an interview).

Re: AI used well can restore middle class jobs [pdf]

#138

Post by bumby »

pydry wrote: post #136
bumby wrote: post #134

Earlier quoted context omitted.

That's exactly the point. To put it in more HN terms, the best-engineered software has no economic value until it is put into use/sold. Created value != economic value. So what is the economic value if its never realized? Nothing. The entire distinction is that there is a difference between types of value; my post focuses on economic value because that is what's related to someone's pay. If you could quantify fundame…

>software has no economic value until it is put into use This is exactly my point. This ^ is categorically false. >If you could quantify fundamental research's value to the economy It seems like you're committing the https://en.wikipedia.org/wiki/McNamara_fallacy

Please explain how shelfware provides economic value. Imagine I make some wizbang software capable of increasing GDP by 50%, but it never gets used. What value did that bring to the economy?

I think you are misunderstanding the fallacy you linked. I've already said multiple times there are multiple dimensions to value. The economy is specifically about goods and services sold, so yes, we should measure production. It says nothing of other dimensions. You can have a job that is of immense social value (e.g., clergy or dog fostering or whatever) while only having marginal economic value. So what other non-productive economic dimensions do you suggest? I suspect anything you name would go into a different category than economics (e.g., quality of life). That's fundamentally misunderstanding my point.

You seem to be saying "See that idle factory over there!? Look at everything it's contributing to the economy!" and I'm saying "It's not contributing anything because it's idle." And there may still be other dimensions to its value. Maybe the architecture has artistic value, or it's history has cultural value. But none of those domains move the needle at all for economic value. I have been very clear in constraining my argument to economic value for a good reason, and it's not about your linked fallacious argument.

Since this discussion concerns wages, my claim was that economic value addition is the most highly correlated with pay. You may say we should compensate people more for what they bring to society rather than just what they bring to the economy, and I would probably agree. But that's not the system we're working under and that is my point.

Re: AI used well can restore middle class jobs [pdf]

#139

Post by 2devnull »

asdff wrote: post #119
2devnull wrote: post #116

Earlier quoted context omitted.

Post hoc ergo propter hoc. Offshoring didn’t increase jobs. That’s absurd.

Higher population and 4% unemployment would suggest jobs have increased since

You don’t know latin I see. Post hoc ergo propter hoc is a well known fallacy where one assumes that something that follows an event in time was caused by that event.

Re: AI used well can restore middle class jobs [pdf]

#140

Post by 2devnull »

pksebben wrote: post #117
2devnull wrote: post #115

Earlier quoted context omitted.

“AI will lower the barrier for entry level position” No. Empirically, has it? It will do the opposite. It removes trust, therefore making existing powers more powerful. Ain’t going to democratize anything but cybercrime.

Not sure I follow you. Do you mean that the existence of AI will undermine trust that applicants know what they're doing? It hasn't really been around long enough for an empirical analysis of effects. It's also incredibly hard to measure effects with the systems themselves being in such flux. For what it's worth, I think categorical statements of any stripe are a little premature. Even the EA folks speak in probabili…

“AI will undermine trust that applicants know what they're doing?”

Yes. Is that not obvious? It’s already a major problem in educational settings. I’ve seen video of people using it for job interviews.

Have you been handed any code yet that was written partially or wholly by gpt? If not, you will soon, and then you’ll get it.

Post reply on HN