Hacker News Reader

Comment by OneGuy123 | Hacker News Reader

Based on how most jobs will get automated some form of an UBI will be required sooner or later.

Now the only question is: how to you phase in UBI while still rewarding extra productive people to keep creating/innovating?

Some will say "those people will innovate out of love". This is not always the case, many people are good at what they do and create many awesome things but given the chance they would rather do something less productive because on some level they don't like what they do that much.

View on HN

Isn't this like saying that the invention of agriculture will mean society has to start feeding hunter gatherers for free? That's not how it went. People find new ways to create value all the time, and yet every time we invent anything new people keep saying "all the jobs are going away." It has yet to happen in spite of the agricultural, industrial, and now tech revolutions.

Which new way of creating value will be immune to automation?

Robots/AI can't even automate many of the existing ways. Examples abound, but: robots/AI can't replace carpenters, plumbers, mechanics, or couriers yet. In knowledge work, they're far from replacing engineers, designers, or authors. They can barely walk on two feet, struggle to open doors and climb stairs, and can't think originally. And we have a long, long ways to go before they can. I'm not worried about "automation" as long as I can clean a toilet and a robot can't.

Maintenance of automation objects, although that's not entirely new it will expand at the same rate as the automation itself, maybe slightly lagging behind.

Improvements to automation; automation on top of automation.

Planetary colonisation planning.

I can see all of those being done better by an AI in the future. Maintenance is already done a lot by machines right now, I don't see that changing.

There is a difference between thinking that many jobs might be automated away and the hidden assumption in your question that all jobs will be automated away. Why do you think all jobs are vulnerable?

Art ?

I think the problem is that humans sell labour that is either physical or mental. Already lost physical work has been automated (see, eg agriculture). So we retreat into mental work, the knowledge economy. But what happens when we have decent AI that can perform most mental tasks?

It doesn’t feel like a continuation of what’s gone before, it feels like a step change. When the machine dominates man in all employable qualities, then it’s game over and capitalism breaks down for most people.

I wish I were wrong, but I can’t see a way out of that. Perhaps it’s good, ultimately, because people may no longer have to work in the future.

An optimistic view I guess is to work out how we transition to a new economic system that both continues to generate wealth and to share it. That sounds very much like a problem that must be solved centrally, which we are not good at, so perhaps therefore (based on historical patterns) we will have to await such an extreme level of capitalism that a revolution forces complete automation hand in hand with sharing wealth, ie coordination will only happen when all alternatives (corruption, brutal autocratic rule, state capitalism, etc) are proven unsustainable.

"Isn't this like saying that the invention of agriculture will mean society has to start feeding hunter gatherers for free?"

How did you make that connection? If you aren't hunting, you can do farm work. If you aren't doing that, you can produce tools. If you aren't doing that, you can go mine ores, make clothes, etc. Its not like agriculture was invented and suddenly anyone who wasnt farming was philosophizing, there is always something productive for you to do. Having a UBI without replacing the people who actually do the productive work is just rewarding those who don't work, you didn't suddenly remove the need for work, someone still has to do all that work that needs to get done, but now there is less incentive for them to do so when they can just kick back and worry about someone else doing it.

A UBI is a socialist fantasy that will only ever work if you have a slave force (read: AI) that does work for free.

"How did you make that connection? If you aren't hunting, you can do farm work. If you aren't doing that, you can produce tools. If you aren't doing that, you can go mine ores, make clothes, etc."

This is exactly my point.

No, your point from what I understand is that even given a UBI, people will find new productive work to do. My point is 1) that with a UBI in place, there is no incentive for people to do any of these new productive jobs, and 2) a UBI by itself doesn't suddenly eliminate the need for the old productive jobs. If all of the old productive work is automated, then yes, I agree that people will find new productive work to do. But a UBI will take away any incentive for anyone to look for new productive endeavors.

I actually wasn't talking about UBI; I was talking more generally about automation eliminating jobs being a shaky premise.

However, I still disagree with you; UBI (as currently proposed) would provide so little income that most would still have to do at least some work to maintain their standards of living. Paying me $1,200 a month would not cause me to suddenly sit on my ass, for example. And, capitalism can still exist alongside UBI: if you want more money, compete in the market to get it.

In France, there is basically UBI, you just need to ask for it. Just having the French passport and no job qualifies you to RSA [1] which gives you about 550 euros a month (according to [1], but nets to ~490 according to a friend finishing law school that had to sign up for it since cov19 is making it very hard for him to find a job right now). I believe you've don't have to have ever worked to get RSA, just being over some age (~26?) and being a French citizen. So you don't have to give UBI to everyone, just the people that need it while they figure it out. Now whether you, as a society, want it is an entirely different story; I've met some peeps that absolutely despise RSA in France.

I think there's also a bunch of programs financing unemployed people so they can go back to school and pick up a more current job once their unemployment benefits end.

[1] https://en.wikipedia.org/wiki/Revenu_de_solidarit%C3%A9_acti...

One thing different about a UBI is you get it whether or not you are employed, so at least it removes the perverse incentive to stay unemployed or lose your 550 euros a month.

That is not remotely like UBI. It has all the pitfalls of a standard welfare program and none of the upsides of UBI.

Does RSA represent the complete package of assistance for someone or are their other programs (food assistance, housing, etc)?

  > Based on how most jobs will get automated some form of an UBI will be required sooner or later.
  > Now the only question is: ...

"Most jobs will get automated" is actually very questionable. It's called "Lump of labour fallacy", although (I suppose) either not everybody subscribes to it, or it's something that tends to go unnoticed.

Fair point, but for example if trucks get fully automated: the USA has 3.5 million truck drivers.

Where are they gonna find a job? My fear is that little by little people will be running out of jobs to find and the middle class will shrink more and more because of this. At which point will the "automation tipping point" happen where the society acknowledges that it's simply not possible to find meaningful employment for everyone?

~70% of Indians are still farmers vs ~2% in the US. Would you recommend India adopt UBI before they discover the tractor and 68% of their country becomes unemployed?

What about the seamstresses destroyed by the invention of the sewing machine?

Forgive my harsh tone - but this is such a false line of thinking "Technology X will destroy job Y" that has been around for literally hundreds of years, that it deserves condemnation as it bares no reality in history or economics.

The self driving truck is no different than the sewing machine.

Past performance is no guarantee of future performance.

The rate of automation is increasing.

Forgive my harsh tone - but this is such a false line of thinking "It's always been this way therefore it will always be this way" that has been around for literally thousands of years, that it deserves condemnation as it bares no reality in history.

To dismiss fear of automation as some resurgence of ludditism is reductive in the extreme. We already see examples of automation killing jobs and seeing workers turned out on their ear because their jobs have been automated.

It's called the Rust Belt for a reason. The manufacturing jobs didn't just skip off to Mexico and China: they were automated away as well. The USA manufactures more now than it did in the 80s, but factory jobs are far fewer.

Most of those factory workers now live off of benefits and Walmart, because their jobs disappeared, driving the income disparity.

The US middle class is dying from automation but you blithely claim "hey at least they're still getting McDonald's jobs".

I think that the consequences of automation should be separated into short and long term.

Definitely, I agree that "automation jumps" do create a short-term problem, if it's not handled by the government. And I think it's a hard problem.

What I disagree with, and think it's a fallacy, it's that it constitutes a long term problem. I stress that long term is beyond the work life of an individual, so the argument "what does person X if they lose their job" doesn't apply, rather, what's the size of the workforce (= employed people).

People who think that the work pool is fixed, typically don't realize that automation (the robots are coming!) is not a modern phenomenon. Automation in a modern sense started centuries ago, so the arguments shouldn't take today as starting point - they should take the 18th century instead; if the argument was true, we would all be unemployed right now.

Nobody can say what what will be the new jobs in the future, as much as 20 years ago, nobody could imagine, say, the position of the SEO expert.

The first automated trucks won't be able to handle everything. There will still need to be a "driver" to ride along with it, at the very least to handle paperwork. Many truck drivers own their own trucks, and I imagine many of these people would be the first adopters of the technology.

The point is that change will still happen gradually, and people will still be necessary.

>I imagine many of these people would be the first adopters of the technology.

Why do you think so? I can’t quite think of a great reason why they would or wouldn’t be the first. My inclination would be that they wouldn’t because 1) self-driving tech would be expensive (more expensive than not having it, certainly), and 2) individuals might not be willing to risk it vs. companies that can afford it to fail and bear the cost.

I think we’d all agree that in the last 20 years, automation has increased in developed economies? So if “most jobs will get automated..”, shouldn’t we see the unemployment rate progressively increasing in those countries? Whereas the experience in the US pre-pandemic argues against this

UBI isn't socialism. The incentive for productivity under UBI is called capitalism, and it means that the people who want to live more than slightly above the poverty level will continue to work, and the people who want to get rich will continue to innovate and start companies. There are plausible (though obviously untested) arguments that UBI will increase innovation by getting people out of bullshit jobs and enabling entrepreneurs to pay their bills until their companies are profitable.

And what about the people who don't want their earnings to be taxed to pay for people who can work but choose not to?

To be fair I don't think UBI is unique in this problem. Even traditional assistance programs struggle with this problem. It would be wonderful if we had some mechanisms that helped people become self-sufficient but that seems to be a very much unsolved problem.

I don't think UBI will be much better at this but the argument that it could be more efficient (i.e. provide the same level of support to more people at the same cost when compared to traditional assistance packages) is what continues to intrigue me about UBI.

"will increase innovation by getting people out of bullshit jobs and enabling entrepreneurs to pay their bills until their companies are profitable."

I think that relies on a definition of "innovation" that is not proper.

What does "innovation" mean?

We already have restaurants. If you open another restaurant have you done anything innovative at all? Most likely not.

You're not going to develop the next generation of microchip in your basement. Multibillion dollar industrial plants are beyond the means of an individual who depends on UBI.

So what kind of new development are you likely to actually create?

Precisely. There is no way I could live he life I want on €1200/month.

UBI is unrestricted income, which makes it so questionable and potential harmful. Regular wellfare with it's rules and conditions can still serve the same purpose, while being better for society on grand scale.

It's true that with more automation we need better wellfare-rules, especially more flexible ones. But no rules at all looks like a strong ignorance against the existing problems if people who are in demand of wellfare now.

There are many who need wellfare because they are simply unable to function on their own or who are not willing to be productive for whatever reason. On the other side there are also those who are good on their own and have the wish to be productive, just don't get the chance.

For the first ones UBI is bad, for the second ones UBI would be good. The question is how to filter them and figure out how to help them both without harming the other side.

Germany already has a system that works well in that regard, even though it's not perfect. But in germany the focus usually is on supporting the people to be useful again, so they cann support themself. Though the problem is this grinds a gear with both sides, creating many complains from certain people.

I think one of the strongest practical arguments in favor of UBI is that the beurocracy needed to decide who deserves to receive welfare is so cumbersome and inneficient that it's more cost effective to not have any restrictions at all. There is a whole subsystem of the courts dedicated purely to dealing with these conflicts. Resolutions are generally very slow and there's tons of overhead s because you have to pay lawyers. Just give every adult citizen a flat amount and you open up the possibility of reducing the size of government in the process of fighting income inequality. This seems like a compromise that could be feasible in the US, but UBI would have to be robust first.

This is sort of along the same lines of the observation that just providing housing to homeless people is cheaper than incarcerating them, and provides better outcomes. Of course there is something about this that people instinctively don't like, and in practice the general population has shown it would rather spend more money to accomplish less in order to prevent someone else from getting something for free.

Indeed, it's a valid argument for those who are proper adults and can handle their live. But it's an invalid argument for those who need support even on that level if living. Bureaucracy is more often build to give those support who need it, even if they don't think so, as they have different goals than society. It's major problem and basically unsolvable.

> This is sort of along the same lines of the observation that just providing housing to homeless people is cheaper than incarcerating them, and provides better outcomes.

Does it?

> Of course there is something about this that people instinctively don't like

It's not just instinct, it's often also experience and understanding things on a greater scale. At least in germany it seems many who work in social services and hotspots are against UBI simply because they see how bad certain people can't handle their live even with all that help given.

If someone is having so much trouble that giving them money or monetary equivalents won't help, then the traditional american welfare apparatus won't be of much help either. It then becomes a question of what kind of living assistance those people need. If they receive nothing, then they end up in prison where we have to pay for them anyway. Most of these people aren't dangerous at all, but don't have anyone to help them, and can't afford mental health treatment. There is alot of overlap here with other endemic problems like our broken healthcare system.

The cost of incarceration in the us is over $30k/person/year, not including exorbitant and exploitative fees levied against inmates and their families for basic communication etc. Just providing housing and staff is much cheaper. The prison system in the US has also become a giant warehouse for these same mentally ill people, whose problems are generally exacerbated by an environment that isn't designed to help them.

> Based on how most jobs will get automated

Will that happen before or after we get fully driverless car everywhere?

> based on how most jobs will be automated ...

Look into the luddites.

There is no evidence to come to such a conclusion.

Reply on news.ycombinator.com