'In the Battle With Robots, Human Workers Are Winning' (sfexaminer.com) 84
Despite warnings that AI will rob humans of jobs, "Somehow we sacks of meat — though prone to exhaustion, distraction, injury and sometimes spectacular error — remain in high demand," writes New York Times columnist Farhad Majoo. AI has yet to replace humans in supposedly at-risk professions like truck driving and fast-food services.
Majoo's conclusion? "Humans have been underestimated." It turns out that we (well, many of us) are really amazing at what we do, and for the foreseeable future we are likely to prove indispensable across a range of industries, especially column-writing. Computers, meanwhile, have been overestimated. Though machines can look indomitable in demonstrations, in the real world A.I. has turned out to be a poorer replacement for humans than its boosters have prophesied.
What's more, the entire project of pitting A.I. against people is beginning to look pretty silly, because the likeliest outcome is what has pretty much always happened when humans acquire new technologies — the technology augments our capabilities rather than replaces us. Is "this time different," as many Cassandras took to warning over the past few years? It's looking like not. Decades from now I suspect we'll have seen that artificial intelligence and people are like peanut butter and jelly: better together.
It was a recent paper by Michael Handel, a sociologist at the Bureau of Labor Statistics, that helped me clarify the picture. Handel has been studying the relationship between technology and jobs for decades, and he's been skeptical of the claim that technology is advancing faster than human workers can adapt to the changes. In the recent analysis, he examined long-term employment trends across more than two dozen job categories that technologists have warned were particularly vulnerable to automation. Among these were financial advisers, translators, lawyers, doctors, fast-food workers, retail workers, truck drivers, journalists and, poetically, computer programmers.
His upshot: Humans are pretty handily winning the job market. Job categories that a few years ago were said to be doomed by A.I. are doing just fine. The data show "little support" for "the idea of a general acceleration of job loss or a structural break with trends pre-dating the A.I. revolution," Handel writes.
Handel notes that despite AI's high performance in analyzing X-rays, the number of (human) radiologists keeps increasing, with worries that the supply of (human) radiologists may not keep up with demand.
One Stanford radiologist recently argued that instead, "The right answer is: Radiologists who use A.I. will replace radiologists who don't."
Majoo's conclusion? "Humans have been underestimated." It turns out that we (well, many of us) are really amazing at what we do, and for the foreseeable future we are likely to prove indispensable across a range of industries, especially column-writing. Computers, meanwhile, have been overestimated. Though machines can look indomitable in demonstrations, in the real world A.I. has turned out to be a poorer replacement for humans than its boosters have prophesied.
What's more, the entire project of pitting A.I. against people is beginning to look pretty silly, because the likeliest outcome is what has pretty much always happened when humans acquire new technologies — the technology augments our capabilities rather than replaces us. Is "this time different," as many Cassandras took to warning over the past few years? It's looking like not. Decades from now I suspect we'll have seen that artificial intelligence and people are like peanut butter and jelly: better together.
It was a recent paper by Michael Handel, a sociologist at the Bureau of Labor Statistics, that helped me clarify the picture. Handel has been studying the relationship between technology and jobs for decades, and he's been skeptical of the claim that technology is advancing faster than human workers can adapt to the changes. In the recent analysis, he examined long-term employment trends across more than two dozen job categories that technologists have warned were particularly vulnerable to automation. Among these were financial advisers, translators, lawyers, doctors, fast-food workers, retail workers, truck drivers, journalists and, poetically, computer programmers.
His upshot: Humans are pretty handily winning the job market. Job categories that a few years ago were said to be doomed by A.I. are doing just fine. The data show "little support" for "the idea of a general acceleration of job loss or a structural break with trends pre-dating the A.I. revolution," Handel writes.
Handel notes that despite AI's high performance in analyzing X-rays, the number of (human) radiologists keeps increasing, with worries that the supply of (human) radiologists may not keep up with demand.
One Stanford radiologist recently argued that instead, "The right answer is: Radiologists who use A.I. will replace radiologists who don't."
How is that "winning"? (Score:5, Insightful)
"Winning" would be if robots did the job and humans had time for more meaningful things than work. The only reason this is misclassified as "winning" (basically a lie) is because it would be, again, just a few that would reap the benefits and many that would drop into poverty. And that is the real problem here.
Re:How is that "winning"? (Score:5, Insightful)
Re:How is that "winning"? (Score:5, Insightful)
Productive, useful work is meaningful. Work that is essentially charity, makework, or bureaucratic red tape meant to reduce productivity and create jobs is demeaning and embarrassing - exactly what would lead to depression (and does) in many people. It does not satisfy the human need to achieve.
That said, the article both hits a point and misses the problem we are encountering. When technology is adopted, it augments us and usually does bring more work. However, we are resisting true adoption of much of the current technology. If we continue to keep the technology at arm's length and resist actual, personal augmentation, we do risk being made obsolete.
As in most such things though, the saving grace will be human diversity. Some nations and peoples will embrace augmentation and obsolete others. The others will then either give in and catch up or fall into obscurity.
Re:How is that "winning"? (Score:5, Insightful)
I would disagree about charity - in my experience those doing it tend to find it extremely meaningful and rewarding. After all, you're generally making a *huge* difference in the lives of those you're helping.
But for the rest, yeah. Few people find flipping burgers, running cash registers, etc. to be meaningful. And those who do some of the most important work for society - custodians, trash collectors, infrastructure maintenance, etc. - tend to be widely disrespected.
Re: How is that "winning"? (Score:3)
I think the PP had a different meaning of charity in mind. Not working to provide true benefits to others. But given a job and income doing useless tasks that produce little or no value to society or a group of individuals.
Re: (Score:2)
Re: (Score:2)
That was the concept of the minimum wage when it was established:
“No business which depends for existence on paying less than living wages to its workers has any right to continue in this country... and by living wages I mean more than a bare subsistence level - I mean the wages of a decent living” -- FDR, the President who signed the first US minimum wage into law
Re: (Score:2)
Re:How is that "winning"? (Score:5, Insightful)
I think you're crediting a lot more social mindedness to people than most have.
I agree that work is meaningful, but most people don't seem to care so much about contributing to society, but to themselves and their immediate social circle. For some a paycheck is enough, and they don't really care about the work. But for others, those who actually want to make a contribution to society? The vast majority of jobs are just that - jobs. Not a Work that provides any emotional fulfillment. Our social machines have become so complicated that being a cog in the machine often no longer provides any emotional connection to the end result.
The people that end up in work-related despair tend to be those stuck working one or more pointless, dead-end, make-work jobs just to survive, with no realistic prospect of improving their situation. Everything from cashiers and warehouse workers to the vast numbers of paperwork wranglers. Even when you do technically contribute a great deal to the larger project, when your job leaves you buried in the internals so that you can't see it, that's likely to be cold comfort.
Re: (Score:2)
people don't seem to care so much about contributing to society, but to themselves and their immediate social circle
That's the result of decades of unfettered Capitalism. It slowly changes society's values to align with it's own: Greed. Why would someone want to make a personal sacrifice when they know that if the tables were turned others would not even entertain the notion of making personal sacrifices for their sake? Or worse, that the law would punish those who did for not being greedy enough?
The vast majority of jobs are just that - jobs. Not a Work that provides any emotional fulfillment.
Again, that's the result of decades of unfettered Capitalism. Work stops being about improving the state of things. As any i
Re: (Score:2)
Not that I disagree, but you're claiming wild speculation as fact, which rarely ends well, and tends to lead to self-delusion.
Most people throughout (90+%) history were poor farmers working to profit their Lord, and feed their families.
Aside from a brief window during American colonialism, which shows no obvious change in common greed levels, you have to get into prehistorical hunter-gatherers to get away from that trend.
And that far back you're dealing almost entirely with speculation from scant evidence r
Re: (Score:2)
many humans will kill themselves in despair.
So it's a self solving problem then.
More jobs and resources left for the winners.
If communities work that way they'd be very unpleasant for most members. There's no winning there.
Re: (Score:2)
Meaningful work is profitable for everyone.
Re: (Score:2)
If we have no useful function in society, many humans will kill themselves in despair.
No, many humans will kill themselves as a result of chronic depression caused by their fellow humans claiming they are worthless for their lack of suffering under a workload the others, on an individual basis, decree as sufficient.
Or to put more simply: Many humans die as a result of bullying from those wanting others to suffer as much or more than themselves. Which is a real fucked up place for society as a whole to be. Especially with increasing automation, in which the requirements for gainful employm
Re: (Score:2)
That all seems... pretty fantastic. You're reminding of a a short story, "A Nice Morning Drive" set in the far future of 1982 where safety standards have tragically made cars too safe, so owners of the new super safe cars roam the roads looking for drivers of older, less safe cars so they can murder them in car accidents. It's a bit of a ridiculous premise, just like your premise that people with no need for jobs will necessarily turn to substance abuse or become compulsive thrill-seeking criminals.
The subs
Re: (Score:2)
The counter to your argument is that air is free. Technically, you have to do some work to get it, called breathing, but air is a free resource. At present no one charges or taxes you for it. For some people (fewer and fewer) water is similarly free. They might have to pump it or fetch it from a stream or collect it falling from the sky, but it's free. That's not true for a lot of people these days. They have to buy it one way or another. Sunlight is normally free. Once upon a time, food was free. Sure, you
Re: (Score:2)
Work is meaningful to humans. Very meaningful. If we have no useful function in society, many humans will kill themselves in despair.
That's a very traditional and familiar attitude. Lots of people in the UK in the mid-1800s felt that way around the middle of 1800s, during the potato famine in Ireland. The consequences of that attitude basically ended up killing probably tens of thousands or more Irish people through the work house system. Admittedly, it's hard to be entirely clear on which people just had a misguided notion about the moral value of work and those who really just wanted to commit genocide against the Irish. Either way, th
Re: (Score:1)
If we have no useful function in society, many humans will kill themselves in despair.
Yet millions of retirees are happily enjoying their lives with no thoughts of suicide.
Re: (Score:1, Insightful)
"Winning" would be if robots did the job and humans had time for more meaningful things than work.
I see pretty much zero evidence that people who don't want to work have any intention of doing "more meaningful things than work".
Re: (Score:1)
I see pretty much zero evidence that anybody gives a fuck about your opinion of what's meaningful.
Re:How is that "winning"? (Score:4, Informative)
That is because you do not understand the statement. For these people work has zero meaning, so _anything_ else would be more meaningful. In fact, this is the easy and obvious case.
Re: (Score:2)
> I see pretty much zero evidence that people who don't want to work have any intention of doing "more meaningful things than work".
For analogy, if Johnny or Mary want to be part of a group but get rejected over and over they might end up alone and behaving as if they don't want or have any intention of joining a group. Doesn't mean they don't want to.
Re: (Score:2)
Not very clear.
I'm also having trouble understanding what the parent meant.
Re: How is that "winning"? (Score:4, Insightful)
Most retirees have no intention of going back to work.
I doubt they consider their time to not be meaningful.
John Henry (Score:2)
"Winning" would be if robots did the job and humans had time for more meaningful things than work.
You make a good point. John Henry [wikipedia.org] "won" too.
Re: (Score:3)
but the society today is exactly as you describe it. Most of the things that you use / buy on daily basis are done by someone else, partially or fully automated. You are given extraordinary amount of time to do things other than growing food, finding and generating energy, figuring out how to heal yourself, making clothes, transporting yourself, entertaining yourself, this and more, much more is done for you in constant basis by people and machines other than yourself. You would have spent 100% of your
Re: (Score:2)
Of course if we actually lived in a world where we had cracked the mystery of how 'cognition' (the ability to THINK) actually worked, could make actual general
Re: How is that "winning"? (Score:3)
That's the major failing of automation coupled with capitalism. Now it's just the owners of the robots vs that owners of the slaves.
How would widespread automation be perceived in a predominantly communist culture? And I mean actually all people working for the good of the whole, not Chinese/Russian communism.
I'd argue that people could still find purpose (ie, important 'work'), whether it was the creation of art, works of charity, or simply repairing the machines that allow individual humans more freedom
Well,l of course (Score:4, Interesting)
You don't have to invest a large amount of your capital into a human. If the human becomes damaged, you can offload repairing it to society, same if it gets broken beyond repair.
You only rent them, you don't have to buy them and front the cost.
Not the Reason (Score:4, Interesting)
Re: (Score:2)
Yeah. For the moment robots are really expensive (though automated software systems can be much cheaper, are technically robots doing intellectual work), and require experts to program them to do exactly what you want - which tends to require much greater expertise in the field than actually doing the job, since you need to program in all the exact decision making trees for all possible situations, rather than being able to rely on common sense, intuition, etc.
If however Tesla, or a competitor, can deliver
Re: (Score:3)
That still depends on how long you need it, how fault-prone it is and how high the mainenance is. In all these regards, a human is superior. You can get rid of it if you don't need anymore (without any worries about environmental issues for just tossing it onto the curb), if it is faulty, just toss it and rent a new one and you're not responsible for a human's maintenance.
Why the hell do you think we did away with slavery in the first place?
Re: (Score:3)
Re: (Score:3)
Let's be honest here, it's simply cheaper and less risky to rent than to own people. Owning people means that you have to shoulder the risk that they get sick, die or are otherwise incapable of performing. When you rent them, that burden is shifted to the person, you just toss them away and rent a different human.
Re: (Score:2)
It does, but all those things will tend to improve with time.
Just to keep things in perspective, if the robot is as productive as three minimum wage stoners (to cover the same 24 hour shifts), then at $20,000 you could buy a brand new one every few months and it will still be cheaper.
I guarantee you slavery wasn't abolished because hiring people was cheaper. If that was the motive, the US wouldn't have had to fight a devastating civil war over it. Providing someone with a minimum of food and shelter, and
Re: (Score:2)
Very few people will be willing to voluntarily work for someone who only pays then enough to afford a couple cups of beans and rice per day, and enough rent to share in a drafty room with a dozen other people.
That mostly depends on how far the next border to some dirt poor country is.
Re: (Score:2)
Not really - even in a dirt-poor country bare survival is rarely too difficult - anyone who can manage to leave is doing better than that by definition. Unless there's a vast "railroad" of people donating food and supplies to make the journey it's just not possible for someone on the verge of starvation.
Consider - a pound of beans (~$1) will feed a person for a couple days, maybe just one if they're working hard, and a drafty shed and outhouse shared by a dozen people costs only a few hundred bucks to build
Re: (Score:1)
You don't have to invest a large amount of your capital into a human. If the human becomes damaged, you can offload repairing it to society, same if it gets broken beyond repair.
You only rent them, you don't have to buy them and front the cost.
Just because your job doesn't give you Medical, Dental and Life insurance and a matching 401k doesn't mean the rest of us are without.
Maybe your "company" just doesn't like to invest in *YOU*.
Re: (Score:1)
You can't lean on phrases like "the rest of us" if they aren't statically dominant.
Exceptions don't make the rule. Someday you might see the irony in your post squawking about how a single data point does not represent the collective. Someday you might see the projection of your self-concerned worldview on a post that wasn't talking about its own personal life in the first place.
The reality remains, proles are used as disposables.
Re: (Score:3)
My company just invested about 10k into me, it's not exactly my problem.
But I'm one of the privileged few. This is by far not the norm. What's normal is to rent a human, use it as long as it's convenient, then throw it away and replace it with a new model. Much like you do when leasing a car.
Hardly a surprise to me (Score:3)
It's been clear for a long, long time to me that the main product of the AI industry is marketing hype and stock certificates.
Truck drivers? (Score:2)
The long term pattern of trucks being driven for vast distances by AI has begun to be established yet, yet that is an example of an industry that will be radically reshaped. So how does the author claim that truck driver's being challenged by AI isn't an issue?
Re: (Score:2)
Totally agree. Loblaw's in Canada is using driverless trucks for moving groceries from warehouse to the store right now: the writing is on the wall.
Re: (Score:2)
> Loblaw's in Canada is using driverless trucks for moving groceries from warehouse to the store right now
When the conditions are acceptable, a small truck from one warehouse, over fixed routes to five selected retail stores.
They just received a permit to operate without an onboard safety driver, which is probably what's behind the publicity push, but it's still not clear if or when they'll be doing away the safety driver or be relying on an operator monitoring the vehicle remotely.
In the short term it l
Re: (Score:2)
Yup, They can't make it deal with smaller streets yet, so it's mostly for highway-reachable destinations at this point. I think before we see truscks doing city-street deliveries, we'll see large transport trucks doing highway long distances.
And it's likely not very far off. There's a bunch of companies working on it right now.
That name rang a bell... (Score:5, Informative)
Re: (Score:2)
How dare technology change and people reassess their thinking.
It didn't fundamentally change, that's why robots still aren't dominating the labor force even in cases where robots could theoretically do the job. In practice, they might be capable of performing a task under many conditions and yet still completely shit the bed in others which are not meaningfully different to a human.
Re: (Score:2)
A bit of useful propaganda (Score:5, Interesting)
All things in time. (Score:2)
in the real world A.I. has turned out to be a poorer replacement for humans than its boosters have prophesied.
Considering AI is still in it's infancy, it's waaay too early to declare human labor victorious. I feel like this is a comparison between vacuum tube computers and the human mind for doing math calculations. Just FYI, the AlphaFold AI cracked protein folding this year which was a problem that had gone unsolved for decades.
Give it a few decades and we'll see just how many humans are left doing menial jobs.
Re: (Score:3)
If it's going to take a few decades then people and society will have time to adjust.
The claim by the histrionic media was that the change would be much faster, with a few years, certainly within one generation. And now it's not.
People can improvise, AI can't. People can extrapolate, AI can, but the non-linear nature of the weights between the layers of the neural net means the results are exceedingly unpredictable.
My Ph.D. Dissertation was on this back in 1997. I'm somewhat surprised that the problem has n
Re: (Score:2)
If it's going to take a few decades then people and society will have time to adjust.
Unlikely because when it becomes good enough at X then all the people doing X will be laid off over a very short period of time.
Posting to reverse bad mod (Score:2)
Proving that while humans suck, AI sucks worse (Score:2)
For now. AI is getting better. It's still best at narrow domain tasks where the major barrier is computational repetition, however, we're applying this approach to improve our cognitive tools (https://www.eurekalert.org/news-releases/965836) quite successfully.
Who could have imagined the process would be slow? (Score:2)
Sheesh. But anyway, I think part of the equation is that it's to be expected that robots wouldn't be an immediate fit, or even a good one, for doing a job that which involved wrapping the controls around a human body. The cab of a truck was designed for a person.
Truly replacing a human truck driver would involve redesigning the truck. The front would be housing for sensors and the circuitry of the AI.
Say, didn't the reboot of BSG have an episode where Katee Sackhoff/Starbuck piloted an advanced Cylon ship t
It really depends on how we define this (Score:5, Insightful)
My go-to example of a job that's left a massive amount of unemployment in its wake is finance.
In the 1970s, how did retail work? Many people paid with cash, some with checks, others paid with credit cards. Someone had to count the cash drawers, add up the checks, and consolidate the credit card slips from the knucklebusters. This then had to be brought to the bank, and someone there had to validate and alter account balances. The resulting statement went to an accountant, who then had to factor in all the different variables and provide a separate set of charts indicating profits and losses, calculate sales taxes, and so forth. While a single Main Street small business may not have hired one person individually for each of these tasks, the man-hours spent probably did add up to about a single full time employee, give-or-take.
How is that done today? Lots of places don't take checks at point-of-sale anymore, and credit cards are almost entirely automated. Point-of-sale software automatically updates the general ledger in Quickbooks, which then generates the financial statements automatically, sales tax calculations are relatively trivial and payable online...businesses still need accountants, but only for a few hours a month instead of a few hours per day. On the flip side, a single accountant can handle 200 accounts fairly easily; a firm of a dozen CPAs can handle thousands (or hundreds of much larger accounts that have more complex ledgers); 50 years ago, more than a few dozen would have been impractical.
Money management hasn't gone away, but the amount of the process that has been automated is so great that it employs a fraction of the people it once did. There's still a need for accountants, but far fewer of them and the lower level finance functions are reduced to the point that it doesn't justify having an employee to handle those things anymore.
THAT is what automation looks like. Honorable mention, newer factories are frequently of the "lights-out" variety; a dozen people to oversee and maintain the machines now handle the tasks which used to employ hundreds. Automation doesn't look like Rosie The Robot, it looks like a shell script.
Even if there are jobs that technology can't automate effectively (social services being a good example), that's now going to end up becoming a saturated market as everyone who needs a job and *can* pivot to such a job will do so.
This turns into a "no good solutions" situation. "The Government" only works if such a solution can withstand being at the whims of your least-preferred politician. "The individual" only works if we accept ruthless levels of Darwinism. "The local community" works if we're willing to spend multiple generations reverting to such a system, keeping in mind that cities aren't going to work well as cities necessarily require food imports. "Tax automation" only goes so far (see "The Government"), not the least of which is how one clearly defines 'automation' to not include cron jobs. "Ban automation" isn't going to work either, as many sections of the economy rely on it; we wouldn't have billions of doses of the Covid vaccine in arms worldwide if every vial had to be handmade.
Short term vs long term (Score:2)
JOHN HENRY (Score:1)
Re: JOHN HENRY (Score:2)
You realize that John Henry doesn't live through that story, right? It's probably the bleakest outlook on automation. Work yourself to death to barely compete a machine, which will be fixed the next day.
Hospitals have perverse incentives to use humans (Score:1)
This is objectively false (Score:2, Informative)
Yes, there's still work to be done, but the issue is that automation is hollowing out the middle. There's a ton of very low paying jobs driving vehicles and cooking food. And there's jobs at the top that require advanced skills relatively few have. But the middle is dying.
This makes sense. If you're automating you go after the middle first. The bottom is cheap enough it's not high priority. The top is too hard.
The problem is it means very few of the kind of middle class jobs needed to ke
Re: (Score:2)
Flexibility versus single task (Score:2)
Even for those repetitive task, you have to look at upcost front and maintenance versus human cost. And robot are 24/7 human not.
So it is no surprise that some task easier to give to robot, and some less.
win the battle, lose the war (Score:2)
In the long run robots will win. A few early victories isn't going to decide the final outcome.
Winning? (Score:2)
Tiger Blood?
The robots will just ... (Score:2)
Not so much underestimating humans. (Score:3)
Radiologists will be replaced (Score:2)
Radiologists who use A.I. will replace radiologists who don't.
160 radiologists will be replaced by 60 radiologists with AI. The increased efficiency means x-rays/CT/PET can be used more times (limited by the bottleneck of building more clinics). After an initial increase in demand, the left-over radiologists won't be able to migrate to a new industry: This isn't a car industry replacing the buggy and whip industry, creating jobs and improving quality of life as side-effects. AI like computers themselves, reduce the number of humans required to complete a job. Th
Why is it all or nothing (Score:2)
I don't see robotics as an absolute threat to a 'job'. I prefer to think of the job as something a human has, and the work involved is the quantity of effort involved. The robots help reduce the latter without replacing the former. In my ideal workplace, the sentiment is 'you show up and push a button, then go home.' obviously not completely literal if you want to WFH or for things that necessitate communication between other meatbags.
If this isn't the goal, then having robots replace us entirely would be q
entrepreneur job (Score:2)
wanted.
invent anti gravity
Cheapness (Score:2)
If we're drawing battle lines . . . (Score:2)
I prefer my robot co-workers:
They don't make a mess in the bathroom or kitchens
I rarely have to re-do something completely if something goes wrong
They don't hide it if something goes wrong
They coordinate with everyone else better
I think Florence (the royal blue one with the red stripe) likes me