Hacker News Reader

Comment by dzdt | Hacker News Reader

I don't buy the historical argument at all: "in the past technology improvements haven't reduced employment, so there is nothing to fear."

The analogy is humans now are to AI as 19th century horses were to the internal combustion engine. For centuries technological advance just increased the demand for horsepower. But when a better, cheaper substitute was developed, demand for horse work dropped rapidly and never recovered.

As soon as there is an AI that can perform better and cheaper than unskilled human labor, the change will come fast.

View on HN

Additionally, unemployment wasn't increased long term. In the short term, however, unemployment would rise drastically and suddenly and destroy people's lives.

The Luddite movement for instance wasn't a group of people against technology in general. They we're against being put out on the street to starve while factory owners got rich because new machines were invented. That sort of rioting and destruction can easily happen again if we take a "tough shit" attitude to people who lose their jobs en masse to technological improvements

That sort of rioting and destruction (namely, the election of Donald Trump) did happen again because we took a "tough shit" attitude to people who lost their jobs en masse to competition from free trade.

Unfortunately, I don't think the lesson was learned, because people mostly just demonize the displaced workers by calling them racists, bigots, and nazis so that they don't have to confront the fact that the magnifying lens of creative destruction might fry them next -- and their families and friends and children and town and any adjacent industry they might attempt to flee to.

On one hand, that's such a terrifying prospect that it's hard to blame people for wanting to avoid the thought of it, but on the other, we really, really need to find a way to address this, because it's a recurring theme that at best won't go away and at worst... well, I don't want to think about that either.

>people who lost their jobs en masse to competition from free trade.

Globalization wasn't really why they lost their jobs. The old jobs left, but they could have had new jobs if we had something resembling an industrial strategy and a progressive monetary policy.

We underinvested in literally every job-creating public good from education to infrastructure since the Reagan era. We decided we could make up for it by putting working people in debt and slashing taxes on unearned (investment) income until it asymptotically approaches zero. Any time the economy gets hot enough that wages might rise, the Fed starts screeching about inflation and hits the brakes on the economy. We've gone out of our way to make sure that people who need to work for a living be as desperate as possible.

An "industrial strategy" usually amounts to the government picking winners and then subsidizing them using my tax money. Governments in general have a terrible record of predicting which industries and technologies have potential. No thanks.

A better approach would be to simply establish a favorable business climate and leave it up to the free market. Help the displaced workers with income redistribution and subsidizing training.

We've been slashing taxes on businesses and the wealthy since the 80s, when exactly is it going to start trickling down?

https://www.theatlantic.com/business/archive/2015/02/why-the...

This approach has never answered the question, "What happens to citizens businesses don't value"

Do we just let citizens in our society starve because rich people didn't need them?

> Globalization wasn't really why they lost their jobs

It's entirely incidental to the subject at hand, but I disagree:

https://qz.com/1269172/the-epic-mistake-about-manufacturing-...

> We underinvested in literally every job-creating public good from education to infrastructure

Absolutely true, and there are massive marginal gains to be had by increasing our investments in education and infrastructure at this moment, but it's important to acknowledge that in the limit of appropriately high funding, education does not guarantee the existence of satisfactory jobs for everyone.

Right on the money with the 'not globalization'.

But the part about conjuring up new jobs out of thin air - that's probably never gonna happen.

With advancing automation, there will never again be as much demand for 'labor' as historically. We've been in a bubble where industrialization still had a need for somebody at the switch. But that's ending, and for good.

In the easily-forseen future, there will be diminishing use for labor. All those with only labor to bring to the marketplace (i.e. most of us without access to capital or goods), there will be no place in our capitalist system as it stands.

> Right on the money with the 'not globalization'.

You sure? Globalization vs robots is entirely incidental to the fact that jobs regularly disappear en masse and we deal very poorly with that as a society, but I disagree that it was obviously robots and not globalization:

https://qz.com/1269172/the-epic-mistake-about-manufacturing-...

Globalization is a step in the direction of automation. Robots are expensive; the scale of the operation has to be large to mitigate that. So globalization is a step in the direction of automation - labor is cheaper in India or China.

Witness: factories come back to America, but only once scale increases (or automation gets cheaper) to the point robots are even cheaper than India or China. The returned factories have no labor jobs - its all robots.

So ok, its all the same effect - nobody is going to pay a living wage for labor any more.

When I was a kid every educated person was convinced that the world would suffer from incredible starvation due to overpopulation by the year 2000. Now I would say the majority of people on HN think that general AI will replace almost all work in our lifetimes.

Maybe. Progress in AI has been fast recently, but it has not always progressed at a constant pace. And meanwhile in the US we have very high employment and a national shortage of truck drivers.

> demand for horse work dropped rapidly and never recovered.

I don't think horse work is a problem. If anything, now they can enjoy their day wondering large vast fields in the summer , and eating high quality alfalfa hay made by a tractor in the winter. And since the invention of the Horse Trailer, they can travel with their owner on interstate highways at speeds as much as 55mph while eating the same high quality hay going to another field.

If anything, the horses benefited the most from the technology improvement.

Mostly, they're kept as pets, and the population has drastically reduced because they no longer have much utility outside entertainment.

That's not exactly a scenario I would call an improvement for human beings.

US horse population is down about 64% since its 1915 peak. Will it be different for people? I guess probably we won't be sent to the glue factory, but that is small consolation.

“As soon as there is an AI” which could be next week, in 10 years, 100 years or never.

Tbh I don’t think it will ever be cheaper to build a robot to do simple manual labour tasks. More likely we will incrementally create environments that require less humans to have the same level of productivity.

>...The analogy is humans now are to AI as 19th century horses were to the internal combustion engine.

While this time things might be different, the analogy people always make in on-line discussions about the jobs for horses is silly. Horses do not have agency. In economic terms, horses were capital, not labor.

That seems like a distinction without a difference. How exactly does a human's agency help their ability to negotiate for a job against a better cheaper robotic worker?

>...How exactly does a human's agency help their ability to negotiate for a job against a better cheaper robotic worker?

Their old job might be gone, but a worker can develop skills that make them more employable - that might be learning about new software, taking classes, getting certified in a new area, etc. A worker can move to a different location where jobs are more prevalent, etc. Automation won’t force all jobs in all industries in all areas of the country to displace workers over night.

As technology changes, economists would say that new job categories tend to be created. While it might be true that this time “things are different” and the job growth will be much slower than in previous times of technological change, that is a very different argument to make than simply using the horse analogy. Are you arguing that a worker is as helpless as a horse when they lose their job?

Reply on news.ycombinator.com