Hacker News new | past | comments | ask | show | jobs | submit login
The idea of a 'robot tax' is gaining steam (businessinsider.com)
36 points by elorant 8 days ago | hide | past | web | favorite | 65 comments

What are the goals here? Is it to maximize the production of the stuff of our lives so that it is available at the highest quality and lowest price possible, requiring the least amount of human toil? (It helps to think about where shoes, houses, food, entertainment, clothing, medicine and all the rest comes from.)

Or is the goal to inhibit change, so that people don't have to apply their minds to figure out how and where they fit into a rapidly changing (improving?) world?

Let's leave aside issues of class and envy, because I think that's a related, but secondary issue that is (at least in part) resolved by the answer to the above.

I wonder if the goal is instead, simply to gather taxes from a broad base of society, from those people and companies most able to afford them, so that all pay towards the cost of society.

As these companies are still relying on governments and the society around them, but aren't paying the taxes associated with employees, we need other ways of getting money from them (and taxing profits is an increasingly bad way of getting money from companies, as they just arrange to have no profits).

>Or is the goal to inhibit change, so that people don't have to apply their minds to figure out how and where they fit into a rapidly changing (improving?) world?

Adaptability is dependent on education and wealth and opportunity, so ideally this tax goes towards helping people who are least able to adapt to sudden automation and loss of income by giving them education and income. We should control change that has a high human livelihood cost so that we do not leave anyone under-served. There should be insurance against economic upheavals.

> ideally this tax goes towards helping people who are least able to adapt to sudden automation and loss of income by giving them education and income.

What is your estimate of the probability that that will happen contingent on a tax being enacted.

Very low in America. We already don't pay teachers enough. Even if a large company with deep pockets (e.g. Google) were to write a check to a state's education budget (so no other government department took a slice of the tax), the increases would more than likely be caught by inflating administrative costs first for no real increases in teaching others.

One of the big goals is to provide time for workers to either retire or retrain for different careers. Right now, a company can buy a robot (capital expense that they depreciate) and the worker is out of job with the possibility of a little severance pay and some unemployment for a short time. A tax that targeted jobs that have been automated away could be used to fund retraining programs and help those displaced workers find new careers.

Why not go the other way and encourage changes in society that reduce the cost of living?

For example, here in California there is a severe housing shortage, and it is widely agreed that the difference between us cheaper places are the restrictions placed by our own laws. Supply is therefore artificially limited.

This is on my mind because I am unemployed and living off of my savings. I don't want to be told what to do all day, and I don't want handouts either. I just want to have the option of going my own way, and to get a square deal on the things I buy.

The only way we have for limiting population influx into "preferable" regions is cost.

At the moment we pretty much let people duke it out economically. If you're productive enough (yes, I'm aware of the flaws there) then you get to live in X area.

Without that I'm not sure how any sort ot UBI can work unless you want to have arbitrary barriers around regions. What's the path for a younger version of me to move from my hometown to London in this world in which my work is valueless?

I don't think cost is the main limit on population density in preferable regions. I would think zoning plays a larger role and likely drives cost. I also don't think we necessarily have a real limit on the number of preferable regions other than land area. Most of what makes a location desirable are the amenities that it has access to, we know how to build schools, hospitals, libraries, parks, and other public services. Rather than making people duke it out for access to such things, maybe we should consider making more of those things available to people?

You can build more, but there will still be competition for what's there. Look at Manhattan as an example.

I'm not sure it's even possible to overbuild in an area that's not economically depressed. If London had three times the indoor residential area, a very possible case would be that it's still expensive, but that people previously living in small flats end up with larger ones or terraced houses etc.

It's hard to imagine how the poor are ever going to be able to meaningfully compete with people who have (at a minimum) hundreds of thousands of pounds of capital. The bottom tier of worker will be bid to their limit at almost any size of dwelling I would think.

Yeah, we still have the "top-level" issue of allocating scarce resources, even after automation "de-scarce-ifies" many things.

Right. I don't think that proponents of universal income really 'get' this. They seem to have the idea that people will just accept their lot, forever.

Personally I'd find that utterly soul destroying. Almost no-one is on top, and a good percentage of people are in tiny flats.

The fact that we have stuff like historical conquests and aristocracies kicking about is manageable precisely because some mobility still exists. Just hitting the 'freeze' button on it would be bloody awful.

Someone who is unemployed and has little savings in California might well just end up being shipped to a project or something.

> Right. I don't think that proponents of universal income really 'get' this.

:-) FWIW, I'm a proponent of UBI, and I get it. (But as I said in a sib comment, to me UBI is an alternative to mass unemployability due to rampant automation, it's not about ideology or utopian dreams.)

> They seem to have the idea that people will just accept their lot, forever. Personally I'd find that utterly soul destroying. Almost no-one is on top, and a good percentage of people are in tiny flats.

That's sounds to me like what we have now. I worked on the Google campus as a contractor, and two things (at least) seem relevant to this thread:

First, early on I realized that this was a model for what I called "secular utopia": the Quality of Life of a Googler is IMO the baseline for what a human should expect on Earth. (Given the physical conditions: the Sun and the differential between it and the rest of the sky; 4By-old self-improving self-replicating nanotechnology; and approx. 2.25 acres per human arable land surface; ...modulus psychosis (what most people call politics) and ignorance. We have knowledge to dispel ignorance and tools to cure psychosis. So really, it's just a matter of getting in the right place and putting one's back into it.)

Er, anyway, cheap and healthy food, ready transport, freedom from avoidable BS, etc.: it's a baseline. If we can get the seven or eight billion humans to a Googleplex level of QoL I would be sooooooo happy. All the other BS can be sorted out orthogonally to that simple, easy, cheap baseline (I'm factoring in Permaculture, et. al. Given land, even desert, it is not hard to bootstrap economically to this baseline.)

Second, a lot of people really are happy in lives that might seem to others to be trite or "utterly soul destroying". It's subjective. The Googleplex was the first place in my life I encountered true "jobbers": people who are totally content (as far as I can tell from external cues) to get up every day, grind out the 9-to-5, go home, and do the whole da,n thing all over again the next day.

One of my co-workers had a stroke. It was in the evening, so most of the cubicles were empty, but even so, nobody noticed and nobody said anything. It was ten days before I realized I hadn't seen him and asked our supervisor what had happened to him. But the most disturbing thing to me was, when he got better he came right back to work.

Different strokes for different folks, eh? (Morbid humor, lol.)

I didn't mean to rant at you. I've had a cold and been stuck inside.

I also question the value of retraining programs, because I think that most people are capable of learning new things.

I was a D student in high school, and I didn't get my undergraduate degree until I was 30. I learn new and interesting things every single day. I really don't think I'm especially intelligent, and so if I can do it, why can't others?

The things you can learn for free through Khan Academy and all sorts of other free online learning platforms are pretty incredible. We have a great number of public libraries. Just a month ago, I was able to check out two books on structural engineering from my local public library. None of this stuff cost me anything!

What's the difference between me and other people? If there is one, it's that I have a different philosophical approach than some others in my cohort. I don't believe that learning is something you do for awhile before you live your real life.

EDIT: On the other hand, perhaps I am wrong! I know that I have also benefitted from classroom instruction (once I had made the shift in my mind give learning another shot), and that there are others that have different learning styles. I don't mean to neglect anyone in my rubric here.

Ironically, our school system in the US was largely designed to create passive factory line workers who did not think for themselves or take initiative. You might even say school was designed to create robots out of people who are now going be replaced by more efficient mechanical robots.

I agree that most people have the capacity to learn a new trade but they haven't had to exercise that ability for many years. Retraining programs can't inject new skills into people's brains but they can setup a path for people to learn those skills if they are motivated. For the vast majority of people, being out of work is a huge motivator to do something, we just need to give them a little breathing room with their bills and life so they can focus.

> What are the goals here?

That's really the important thing to nail down, eh? Wendell Berry has a essay titled "What are People For?" (As an aside, while looking for a good link, i came across a site advertising "BUY CHEAP ESSAY Fast & Safe Academic Writing Service" expletive-deleted what a world.)

Anyhoo... the answers to "what to do about rampant automation" fall out naturally once you settle the fundamental issue. FWIW, YMMV.

An interesting aspect of this is that the definition of both "robot" and "automated" is terrifically fluid.

There are all sorts of productivity advancing tools that tread the line.

If I told you I'd built a machine that halved the number of people/time that it took to unload a truck full of boxes you might think that it pretty clearly would fall under this but take a look at the device:


I'm still pitching Buckbee's law: "A robot's utility is inversely proportional to how much it looks like a robot." and in practice, I think you'd find that almost every real gain in productivity doesn't actually look like a "robot".

I'm surprised by the reaction here. It seems fairly intuitive to me. People used to capture part of the value creation with their wages and they are being replaced by robots. That value that used to circulate in the economy now just goes to owners (wealthiest section of society) instead of circulating. This just takes that same value back.

I don't know the best way to implement it, but I think it is necessary if you want to prevent wealth inequality from getting worse, faster over time.

Generally, the most efficient thing to tax is the thing you don't want.

What you are proposing is a good thing (greater efficiency) with a negative side effect (tilting economies towards the rich) but rather than tax the bad thing, you're taxing the good thing in order to reduce its negative side effect.

For whatever reason it seems like taxing robots is more socially acceptable than taxing the wealthy.

You really don’t want a couple million unemployed and unemployable citizens at your doorstep one day angry for why they suddenly can’t afford to live. You don’t want it as a government and you don’t want it as a company/corporation.

Back in the 60s, there was the joke that it'd be less expensive to just bribe the Vietnamese instead of fighting them.

Instead of dropping bombs they should have just dropped cigarettes, whiskey, magazines, food, clothing and a variety of American consumer goods.

> For whatever reason it seems like taxing robots is more socially acceptable than taxing the wealthy.

But these two often overlap. If any actual revolutionising automation by robots is coming, only the wealthiest corporations/people can and will afford it. As your parent poster said, robot tax is just trying to make sure that they won't hold on to their money under the mattress until the end of time. Redistribution of wealth is critically important.

I think there are at least two errors with your reasoning: one is that people don’t have inherent (economic) value that is captured by working for wages. If a robot does my job, why would I feel entitled to capture any value?

The second is the idea that the owners are less likely to circulate money in the economy than the employees. There’s no value to the owner in sitting on that money, so why would they?

First, it's not that they are 'entitled' to the value, it's that our entire economic model is based on continuous consumption. If vast swathes of the population have no way to get money because robots do the work and only the wealthiest capture the value, then it collapses because there is nobody with money to consume the goods produced.

Second, the point you're getting at is trickle down economics. Personally, I don't think it works.

Because robots are not taking the jobs away. Some of the wealthiest, biggest companies in the world have more employees than ever in spite of having the the most resources to invest in automation. Successful businesses tend to add people, not replace them. What automation does is allow people to be more productive but that is not the same as replacing people.

Exactly this. Until robots can entirely replace humans, there will still be the need for humans to do certain jobs. As robots become commoditized, companies will start to compete purely on the pieces that can't be automated, ie the human element and increase demand for human workers

All true, only those people will be probably 1000x less than all working people currently, due to highly specialised and manual and creative skills that not everyone would (or could) acquire. So this still doesn't seem to address the problem of a huge portion of humanity being displaced and left to live under a bridge.

That's not quite true.

If you took a single production line and replaced the workers with robots - then yes. Those workers have lost their jobs.

However if there was a second hypothetical production line that made a product that was now economically viable (or just cheaper) by partially automating it, then that's made more manual jobs available (maybe the line didn't exist before, or now as the product is cheaper, there can be two lines).

So you've still got people working - and you've now also got cheaper/new stuff for them to buy.

Watch making is a good example. Wherever you are on the planet, you can afford a watch - something that used to require incredible wealth. This is due to new tech and automation. Doesn't mean that nobody's handcrafting cogs with little files, just means that those watches are now only being bought by the wealthy.

In the current economical climate that seems to tilt more and more towards huge inequality (which is already a fact, mind you, it's just tilting even more in the same direction) I think we cannot afford to simply cross our fingers saying "oh well, maybe new jobs will magically appear". Yes it happened in the past, sure. That's not guaranteeing it will happen again this time around though.

It'll all work out overall - that's the defining concept of the system.

Whether it'll work out for individuals though - well the system doesn't care (and that again is a defining concept)

There are a lot of robots in agriculture for some time now ( just naming 2 : the tractor and the milking machine ).

What is different now?

Scale and artificial intelligence.

The scale of the automation/robotization of agri has been humongous.

And broad application of AI is nowhere near.

A robot tax is the most ridiculously luddite political action I can think of. How about instead of taxing technological progress we fix the tax system so it actually profits and wealth accumulation.

The only way I can make sense of this is that the robots provide a surrogate buffer for the wealthy to feel "less attacked". Sadly wealthy people are not interested in the greatness of their planet or society anymore. Somehow they think their private enterprises that serve ads are making society great or leaving a legacy.

It's another everything tax. They just thought up a new word for it.

I imagine this will be circumvented in similar novel ways much like the UK's window tax [0] was centuries ago.

Multiple robots will be attached to one robot "brain" and argued that they are all the same robot.

Or perhaps they will go the Volkswagen route and lie to government regulators. Whenever the government bot taxers walk in the room, all of the bots stop moving and act like smart compacting garbage cans.

Or perhaps costs of robots will dwindle so much that they will become disposable. Come bot check day, the majority of the workforce will be unrepairable and destined for the bin.

Or perhaps they're making high end CPUs with high margins that sell for $500 apiece, but the government thinks you're making plastic bags with low margins that sell for $500/pallet.

This will never work imho... or rather, it will never be "fair".

[0] https://en.wikipedia.org/wiki/Window_tax

> At that time, many people in Britain opposed income tax, on principle, because the disclosure of personal income represented an unacceptable governmental intrusion into private matters, and a potential threat to personal liberty.[3] In fact the first permanent British income tax was not introduced until 1842, and the issue remained intensely controversial well into the 20th century.[4]

Oh, how times has changed...

A robot is just another part of capital. Why not just tax capital directly?

Because there's a profoundly negative impact on investment and economic growth. Maybe the 99% don't see as much direct impact from said growth, but it stopping would still hurt them greatly.

As an aside, what you're describing is not a tax. There's neither a transaction nor any exchange of good or service involved. It's a direct confiscation of wealth, currently unconstitutional, and I'd call it theft. It's not skimming off the top of a transaction, it's literally the feds reaching into your pocket and calling that money theirs because enough other people said it was okay. It doesn't magically become not theft because the thief has a badge or a "democratic mandate" (enough others saying that yes, they do indeed want your money).

I'm not sure that the argument about disincentive is any different from taxing income. We want to encourage both wealth and income, and taxing them does discourage them. But the fact that taxing income discourages people from getting income isn't a strong argument against income taxation altogether.

I'm not American, so the unconstitutional argument doesn't mean much to me. But doesn't the USA have property taxation for land? Wouldn't your arguments apply to the property tax in most states? It's nothing more than the government saying you owe money because you own wealth (in the form of land). Are property taxes theft? Unconstitutional?

Absolutely it's different from taxing income; it demands liquidity. I.e. an entrepreneur with a billion dollars in stock has a net worth made mostly of "paper gains". Very little of that is likely liquid. By letting the money continue to work as an investment, it can keep growing and producing returns. Effectively adding liquidity requirements for private investment would be devastating.

> doesn't the USA have property taxation for land

That's state-level, not federal; the constitution gives the feds no power to confiscate capital. More importantly, it's a direct fifth amendment violation:

> nor shall private property be taken for public use, without just compensation.

Not only is there no enumerated power (enough to make it unconstitutional), there's a specific prohibition.

That said, I'm not a huge property tax fan, either. I'd rather the gov't just bill each citizen for services used as an average, at least at the federal level.

I'm sure many people would prefer to. But making policy in a democratic society requires compromise. If it's easier to reach a consensus that this particular kind of capital needs to be taxed, it might make sense to just enact that tax and worry about other kinds of capital later.

But how will that avoid the problem of slowing down the future by sapping investment into it? I know it's a little more complicated than "more money means more X", but that actually is the premise behind every cancer and AIDS walk out there.

Would a washing machine count as a robot?

Depending on how defined, it could easily include: a self-service elevator, 3D printers, CNC mills and lathes, a farm combine, automobile service lifts, farm irrigation, power tools, construction equipment, pneumatic tools, power screwdrivers, a dishwasher, an automated telephone exchange, internal combustion engines [or even steam engines for that matter], intermodal container systems, an IVR, spreadsheets, word processors, microwave ovens, lawn mowers, leaf blowers, hedge trimmers, automated boilers, sous vide circulators, any web-based order lookup system, automated parcel sortation machines, possibly all of e-commerce...

It seems insane to me that we’d seek to tax productivity improvements in order to encourage maximizing the amount of human labor required to accomplish a given outcome.

“Spoons Not Shovels!” https://quoteinvestigator.com/2011/10/10/spoons-shovels/

|It is a terrible idea. I dunno how someone who is considered smart like bill gates would endorse it.

We need this. And computer tax. And car tax. And refrigerator tax.

But not just a one off tax. A recurring tax to capture all the efficiency and job loss the horse and buggy industry , secretaries, ice delivery industry, etc. have incurred.

Maybe fund guaranteed minimum income with taxes on automation.

And doom us to low wage low productivity feudalism. No thanks.

Surely doom us to a life where the robots do the work and we lead interesting lives of leisure

Unfortunately, I increasingly fear we are doomed to a life where robots do the work and 99% people people die under a bridge.

Huh? That would be robot subsidy or high minimum wage or UBI. (UBI works the best, though in the short term high minimum wage might actually have more affect has you get to decrease labor costs and not pay higher taxes on profits.)

> and we lead interesting lives of leisure

We'd like to think this, and maybe some would do "interesting" things, but the fact is that most people turn into couch potatoes. People who retire early tend to die early; stop working and so will your body.

> but the fact is that most people turn into couch potatoes

> stop working and so will your body

Citation badly needed here.

I can tell you that before my peers joined the workforce, everyone was part of sports and clubs and community activities. After they joined the workforce, that stuff largely dropped away, because work is too time consuming.

Well, sure, if you consider 55 as early retirement, I could see how people would be ready to check out and do nothing. These are people that have likely worked full time for the last 30+ years.

What I want to know is what happens at scale if people retire at 30 or 40, or maybe if they never work at all. The idea is that automation might free up people from work entirely. I don't think it's fair to extrapolate data from people who have been working for 30+ years and then had an "early" retirement.

You're right that it's a loose extrapolation at best, but it's rather the best data we have, or at least the best data of which I'm aware. You asked for citations for my claims, and I provided them; I'm showing what happens when people _stop_ working early, not when people never really start or don't for long.

That aside, a society in which no one supports himself but rather is dependent on the noblesse oblige of the state to survive is a dystopia I hope never to see.

> We'd like to think this, and maybe some would do "interesting" things, but the fact is that most people turn into couch potatoes. People who retire early tend to die early; stop working and so will your body.

50/50, dude. I will never stop working -- but I'd like to work not for money. I want to work on what I find interesting, and some of it can actually be very productive for society. The mere coercion to work or you die of hunger or freeze under a bridge is ruining people's desire to do stuff outside of work -- that, and the fact that work is too time- and energy-consuming.

I don't dispute that a lot of people turn into couch potatoes when they don't have to work. But must this doom everybody else who would do interesting and society-benefitting work if given the chance? Is it fair that this chance is taken away because a huge chunk of humanity would just be useless in front of a TV?

I don't think it's fair.

Plus, if we actually have an economy supported mostly by robots, who cares if those people are couch potatoes, really? If that won't cause an economic crisis then why would this phenomena be important?

You're right. While practical implications are a concern, the biggest issue I see is a profound ideological difference: a society in which everything is produced by machines and the average citizen receives everything he needs to survive from the government sounds like a dystopia to me. I don't trust the state to administer this prospective neo-Oceania. Citizens who are abstracted from adversity will become disinterested in politics and apathetic towards the inevitable tyrannical government that will ensue.

> Citizens who are abstracted from adversity will become disinterested in politics and apathetic towards the inevitable tyrannical government that will ensue.

Sure. That's pretty much what's going to happen. But mind you, I am not seeing things being much better now compared to that future. Democracy to me seems to work only on paper, while politicians, lobbyists, corporations et. al. do whatever the hell they want. But I am not willing to derail on this point, just sharing an opinion.

> ...a society in which everything is produced by machines and the average citizen receives everything he needs to survive from the government sounds like a dystopia to me.

As with everything else, it's a balance. I'd like to not think about utilities and rent and basic food needs, yes; on the other hand, I am okay with the idea that if I want that 75" QLED TV, or a car, or the best chair recliner on the market etc., then I'll have to work a bit for those.

One way to push people into working even with UBI in place is likely to make the places where food is distributed for free inconvenient or not very ergonomic; many people would like to avoid certain inconveniences or even forced social interaction so they're likely to work just to avoid changing their habits. And publicly funded places gaining bad entropy is pretty much a given anyways; nobody has to do anything, it just happens eventually, by itself.

> I don't trust the state to administer this prospective neo-Oceania.

Me neither, but not sure if for the same reasons as you. Governments are, in the end, hugely ineffective. That's basically their defining characteristic. The moment they have to do something mildly responsible they are very happy to outsource to the private sector, and things turn to shit in months. Government is only really concerned with gathering taxes. :(


Overall, I believe systems like we are discussing can be made to work. But I am not sure I trust people to do it. And of course, there's going to be a world war before anybody remotely influential will allow a machine to call the shots in their place.

> a so-called "robot tax" faces significant hurdles in design and implementation.

To put it mildly.

I want to make litter-bots: Imagine a ladybug about the size of half a beachball. It crawls around and collects litter. It's mostly autonomous, but to really do a good job it needs a minder, a human "oracle" for navigation and trash?y/N queries, etc. (The minders do not necessarily have to be physically nearby to the 'bots.) ML advancements have drastically reduced the amount of human oversight needed but not eliminated it entirely.

In that context, what should the economic relationship be between the minders, the clients, and the robot factory? (Consider that the robot factory is likely very automated, a robot robot-factory.)

FWIW, I had originally thought I would set up a kind of volunteer distributed remote workforce with a UI that folks would play like a game.

Another possibility would be to sell or lease 'bots to operators, who are then responsible for finding their own contracts. The litter-bots would be akin to landscaper's tools (like an AI leaf-blower or chainsaw.) "Let the robots do the work and we'll take their pay."

I do NOT think it would be useful to for the government to tax my litter-bot-company's robots and then somehow redistribute that to... whom, exactly?

I am in favor of Universal Basic Income, but only when the robots are good enough to seriously impact the average person's ability to compete in the job market and get work. (Like, me. If I wasn't good at computers I'd probably be homeless!) A given economic regime must provide jobs or bread or "you're gonna have a bad time."

I think we're at the tipping point: the hardware is COTS (Common Off-The-Shelf) the software is coming together nicely. There are already robot vacuum cleaners and lawnmowers that work pretty well, etc.

So, yeah, what's the economic program for the automation revolution? Lords and serfs? Or some sort of Marxist distributed ownership of the "means of production"? (Probably not: even the communists didn't actually go through with that in the end, eh?)

Why is Daiso so much more popular than RepRaps?

> I do NOT think it would be useful to for the government to tax my litter-bot-company's robots and then somehow redistribute that to... whom, exactly?

As much as I think a robot tax looks to be one of very few possible ways to improve current situation, I also agree with you that the whole thing isn't well-explained or argued for. It does sound like this: "government vacuums even more money but 10 years later, people still live under bridges or in slums". Sadly, knowing the average politician's thought process, the odds of us ending with an economic inequality shifted from corporations to governments is extremely likely. We the people are greedy no matter what positions we occupy...

> "Let the robots do the work and we'll take their pay."

I like that a lot, actually. It's a much better idea. Robot goes to work in my name sake, I get money wired in my account and I get the final say in what does the robot do and when. The reliability of the transaction hinges on economical interest (I want the robot to work with a stable schedule N hours a day because that makes me money).

Really neat idea.

Cheers! I can't take credit tho. I believe it was Dr. Philo Drummond or the NENSLO who said it first.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact