Instead, it's targeted at replacing graphics artists and software developers...
Manual work has been getting automated away for over a century. Tech has already eroded low and middle class jobs for decades. To think that similar won't happen in knowledge work (or that we shouldn't let it), feels like a very techy type of NIMBYism
I'll concede that automating manual labor has made consumer products cheaper. Whether that is wholly a net benefit, I think, is up for debate. When you consider the externalities like the environmental effects of rampant consumerism and the erosion of the middle class via a reduction in manufacturing jobs, I think there's plenty of basis for at least arguing there is tipping point on automating away manual labor. To say automating manual labor bluntly (and without careful consideration of blowback) always leads to a better society, I think one needs to define what they mean as society and what metrics they're using to define "better." I, personally, don't think raw GDP is a good measure of the health of a society.
If anything, I think automating knowledge work has as at least a potentially larger upside. The downside will largely be borne by the upper-middle and upper-classes, which is why the GGP can read like a classist perspective.
That aside, what exactly is suspect about finding meaning or joy in work? Is the quintessential experience meant to be soulless and grating?
Of course not. Few things in life are more fulfilling than a job you absolutely love and are excited to whack at every single day.
But your question raises another one of equal stature: Is the quintessential experience meant to be a job? If we are creative enough to make rocks that can learn to do our jobs for us, surely we are creative enough to craft an economic model which allows us who no longer need to work to paint or write poetry or rebuild antique engines without needing to starve?
As far as crafting a post-scarcity economic model goes, it's not the problem of dreaming one up, it's the pervasiveness of scarcity. Even if all of humanity's basic necessities are one day a given, scarcity won't disappear, just shift around (maybe as transportation for the otherwise-infinite supply of consumer goods? Or living space away from dense urban centers? Maybe even the kind of heuristic analysis abilities humans are unmatched in to keep the matter replicators functioning?)
More to the point, saying that this kind of luxury will exist in a thousand years doesn't nullify the concerns of the present, and probably wouldn't convince most people to forfeit their employment to machines likely owned by the uppermost classes.
I think there's a distinction in that I was referring to "knowledge work" which isn't explicitly creative/expressive/stimulating. Many would classify much of research or law as "boring" even though they are knowledge work.
>what exactly is suspect about finding meaning or joy in work?
I did not mean it as wrong to find joy in your work. On the contrary, I think that's a worthwhile goal. But the distinction I make is that any work can be found to be fulfilling. It's the distinction between "finding your passion" and "cultivating your passion." I've worked with people who found meaning in their job cleaning offices and others who treated the design of rockets as soul-crushing. I think it has more to do with the person than the job. So I push back a bit on the false dichotomy created by classifying "knowledge" jobs as inherently worth saving from automation while manual work should be fodder for it. I also think the focus on a job for fulfillment is a bit of a red-herring. I think what people really need are to be valued members of society and, for many, a job is a means to that end (and maybe not even a good one).
Higher wages in agriculture would mean a combination of less profit and/or higher consumer prices. That is not a sin. It is mostly politicd that demand that food prices have to be low so that the population don't realize how poor they are and much they are being robbed on corruption and waste.
It's perhaps worth pausing to consider if we have any kind of useful ability to predict blowback on anything but a trivial scale. It was fifteen years between the invention of HTTP and the creation of Facebook. It was twenty five years between the invention of HTTP and the Cambridge Analytica breach. I do not think any person could have meaningfully predicted those on those timescales.
If we cannot currently usefully predict blowback, then what reason is there to try to anticipate it as a matter of policy?
We can certainly predict that when automation happens people will lose jobs. In a country where productivity is king and self-worth is tied to employment, we can predict this may lead to a crisis in society. And we also know people tend to turn to vices like drugs during those times. So I don't think things like the opioid epidemic that hit the former manufacturing centers of the country are unrelated nor completely unpredictable. They just sometimes take decades to play out.
I'm not making a Luddite case that technology needs to be stalled. But I'm also not making the techno-optimist case that everything will eventually work out if we just plow forward with reckless technological abandon. The latter runs the risk of a lot of increased human suffering in the short term at the very least.
You're completely right that it was easy to predict that there would be some consequences to automation and people losing their jobs. Yet I do not think it was easy to predict what shape those would take. As a result, it was functionally impossible to offer useful policy measures. You can say "We should reform society away from believing productivity is king and self-worth is tied to employment", but that's itself not specific enough to be useful. "This may lead to a crisis in society" is similarly rather non-specific. How do you craft policy around "this may lead to a crisis"?
In practice, I see two recurring patterns when people try to predict blowback. First, people use fears of blowback to launder their anxieties. If you look at the conversation around AI, you will see this happening in many forms.
Second, people often use predictions about blowback to advance policies they wanted anyway. Artists want to be hired more and stronger intellectual property laws, the same things they wanted yesterday. Advocates for saving small towns in the rust belt will suggest the same retaining and social safety net policies they suggested yesterday before anyone asked them to predict blowback.
In my opinion, these two patterns are deeply linked. They are both about trying to turn confirmation bias into policy. None of the answers from this are automatically wrong, but none of them are novel. Most worryingly, neither approach offers any kind of way to reliably predict blowback so it can be dealt with via policy.
In my career, I've seen any number of engineering teams devote significant time and effort to trying to solve technical problems that never arose. Not because they were solved in advance, but because the team's predictions about where issues would arise were wildly incorrect. From this, I have drawn the lesson that we are well-advised to approach the task of trying to predict failure in complex systems with deep humility.
The more complex the system, the more humble we need to be. At some point, trying to make any prediction more specific than "something will probably go wrong" becomes a poor use of time.
This is neither the Luddite case nor the techno-optimist case. It's an argument to be skeptical of our own ability to make good predictions about the future except in, as you wisely and correctly say, very general ways.
Let's dilate on "This may lead to a crisis in society" to try to get to a policy. If we can agree on two things we might be able to get a rough scaffolding of a framework to discuss policy. 1) government programs, like everything from social security to roads/bridges take money to run and 2) the vast majority of federal funds come from taxes related to work, like income taxes and social security taxes. By extension, if automation effects jobs, it then affects the programs that create a stable society.
So one aspect is: as automation takes people's jobs, it potentially threatens the ability of government to fund its programs. If a society ignores this, it faces a potential "crisis" if those programs help create the conditions for a stable society. There's a few ways one could address this. On the cost side of the equation, we could use austerity measures to reduce the cost burden. There's certainly something to be gained here, and it would be a long digression to decide which policies are a priority. (For example, I've heard research saying that roads provide the most benefit on a cost basis, followed by early education programs like Head Start). On the supply side of the equation, it seems like there are two options: a) help workers get replacement jobs that pay at, or near, what they had before their job was automated away or b) get the money through a different, non-income based policy. It didn't seem like we did a good job crafting policy in the rust belt related to a). There wasn't much re-investment into those communities or workers, compared to what was gained by automation. There are various ways to address b), including restructuring corporate taxes or instituting an automation tax to make up for the displaced incomes formerly garnered by workers paying a tax. But we went the other way on those, too.
While I concede those are very high-level, the intent is to show there are real discussion points that can be crafted into policy and it's not just some hand-wavy rhetoric.
Automation started in the 1780s, with the industrial revolution. The initial impacts had a lot to do with creating vast numbers of jobs, driving down the cost of all kinds of consumer goods, and heavily driving urbanization. I can't see any easy way to get from there to the rust belt if I'm someone looking forward in 1780. Right now we can treat this as obvious only because we have the benefit of hindsight. I cannot imagine any way in which the modern history of Detroit would have been reasonably and usefully predictable from 1780 (at the time it was a frontier fort under British military control).
You're right, impacts can be decades off. They can even be centuries off. There were a lot of equally credible people who thought automation was going to have utopian consequences that didn't include people losing their productive economic positions. This isn't a binary, either. There were plenty of other possible outcomes as well. How were people in 1780 to know what we do know? What happens if every predicted outcome is taken seriously? What happens if they're then all wrong, or not right on a sufficient timescale? I know how I would expect that to interact with limited government resources.
At the end of it, I think we're likely limited to dealing with consequences and trivially short-term prediction. Those, at least, we have a reasonable shot of observing.
When your system is predicated on taxes borne from income, and you implement a disturbance that displaces that income, it doesn’t take a rocket surgeon to see there will a problem looming on the horizon. It’s really just connecting two very obvious points that doesn’t need hindsight centuries later to see. People have been ringing the alarm for decades now, policy makers just found it easy and expedient to ignore them.
I think many politicians knew this but policy development has a different wrinkle. Politicians are incentivized to take a short term perspective because election cycles don’t operate on the same time scale as large scale societal issues. It’s the same reason they are incentivized to raid the general coffers for projects that will be completed during their tenure and leave long time horizon problems for someone else to fix. (If you disagree, look at Illinois and their public pension issues)
(And not to be pedantic, but automation dates back a couple thousand years, at least to the use of water wheels to automate grain flour production. The industrial revolution dates to the late 18th century. While they are related, they aren’t the same thing)
Replacing a task with AI that otherwise will be done chemically with environmental degradation or not done at all because people won't or can't-- AI in that context makes things better.
As for attempts to deploy AI as a replacement for sophisticated human work like "automating knowledge work", the output will be worse and it's a race to the bottom on quality. As we see from our algorithmic-driven tech industry friends, robot-led workstreams benefit society only in the short run. In the long run, people come to miss-- and their lives are diminished by-- an absence of quality, humanity, agency, caring, fairness, all the things algorithms cannot and will not ever be or do.
Anyone who is serious about this topic only discusses in terms of AI + humans (as a combination) and not AI as a replacement for sophisticated work. Statements beyond that reflect a lack of appreciation for either the complexity of the topic or for what society actually expects out of a "knowledge" workstream.
>an absence of quality, humanity, agency, caring, fairness, all the things algorithms cannot and will not ever be or do.
I would argue that automation can be quite better than humans on many of these domains. When they do fall short, it's often because they are reflecting (and in worse cases, amplifying) the shortcomings of humans in these same areas.
Can you be explicit?
With LLMs I use it as a tool to improve my efficiency at translating thought to code, that's good.
With image generation, I struggle to see the value, especially relative to its harm even apart from taking jobs from artists. What I see is:
* The increase in possibility of fake photos/fake news (the recent photoshopped Kate Middleton comes to mind)
* A new level of degenerate coomer addiction from AI booba
* Mass production of CSAM (I believe this to be a hugely underreported issue from what I've seen on reddit/civitai)
Diffusion models/image transformers are incredible and fun technological marvels but as far as I can tell current usage is at best a fun toy.
Same for writers, another friend is writing a D&D campaign. Again, he doesn't have the skill himself to do good digital art, and dozens of custom pieces is way too much cash at this point. But some carefully-done AI pieces, they mon't be perfect, but a damn bit better than nothing!
There's an argument that neither do most jobs [1], but that doesn't mean they are unaffected by automation.
[1] Graeber, D. and Lou, L.I.T., 2019. Bullshit Jobs: A Conversation with David Graeber. Made in China Journal, (2).
Agreed, lets not add to that pile of useless jobs
I'm all for getting rid of bull$hit jobs to allow people to pursue meaningful, creative work. And while a lot of people laude the automation aspect, I don't see many proposing ideas on how to close that gap.
A lot of that can already be automated. Push the trash into a series of shredders, until it's powderized, then use magnets to pull out metals, and differentiate the rest by weight, IR spectrum and air streams. Then, one can use a spectrum of solvents to target plastic polymers to get back the precursors.
Unfortunately, for plastic polymers that's still way more expensive than to just dump the wholesale trash on some Asian or African piece of land where it will eventually either degrade into microplastics or be burnt by extremely poor people who hope to get some of the metals back.
It would be nice if we would not stigmatize manual labor. There's nothing wrong with manual labor. For some people, it's ideal.
In a city where I used to live, I'd frequently come home on red eye flights. Over time I started to recognize a couple of older gentlemen mopping the floors in the early morning hours.
Then one day, they were gone. Instead, there was a slow-moving robot mopping the floors. One clearly too large to get into all the places where the dirt accumulates, but... Automation!
A couple of years later, I learned that the airport was part of a municipal program employing intellectually challenged adults, and that the cleaning guys I saw were probably part of that.
Now when I travel through that airport, I always wonder what happened to them. Did they find another mopping job? Are they homeless now? Do they have to live under a bridge because society fetishizes technology and robots? Because budget spreadsheets always have a line for "money saved," but never a line for "lives improved?"
Mopping the floors at the airport is not a great job for a tech bro who wants to make seven figures. But it's a good, honest job for a lot of people.
And to describe manual labor as "abuse" is, frankly, disgusting.
Perhaps in a different world manual labour could be something good, but i dont think its honest of you to say that you think people who do manual labour are not being abused.
The next time you drive through a construction zone, let the construction workers know how you feel. I'm interested to hear their responses.
EDIT: Since HN doesn't permit me to respond to your response, I'll note here that I have worked plenty of manual labor jobs. I've been poor. Very poor. I'm better off today than I was then. But, I don't recall ever thinking that having a job involving manual labor was abuse. It was just a job. It's what you did to pay the rent, and hopefully buy some ramen and canned potatoes for the week.
I've done landscaping, and mucked out toilets, and cleaned out the pits of elevator shafts, hauled endless boxes from one end of a factory to the other, and I've always taken pride in my work. Perhaps I was just brought up differently than you were.
Perhaps you can try working at a mcdonalds or something or as a janitor for a while, see how you feel about your employers or the institution that is employing you.
Regarding ethics of work, you are talking about something completely different. Being proud of working an honest job is not the subject at hand.
I enjoy manual labor and I wish it was easier to get paid for it. I find cognitive work more pleasant and satisfying when I devote part of my time to external physical effort, which forces me to focus my mind outward instead of on abstractions.
It seems to me (reading your other comments) that your issue is not with manual labor per se but the capitalistic system that devalues it and gives many manual laborers as little autonomy as possible.
Art, writing, and basic programming are just the easiest ones to tackle first. You have fast iteration cycles because iteration is virtual.
This temporary period of difficulty is the price we pay for AGI, which will yield far more benefit in the long run.
This is questionable at best. Im glad you have optimism, but do you have any facts that show it won't be a long run benefit like cfcs, or lead in everything?
> This temporary period of difficulty is the price we pay...
Im glad you're signing me up for difficulty. I get that you probably think you stand to benefit in some way from AGI, so why don't you pay the price yourself instead of volunteering others? Maybe it's because the difficulty that it causes really really sucks, and the benefit isn't worth paying the price yourself?
It's irrelevant what the average person thinks about AGI because they can't see past the short term fear of losing their job.
They can't see the billions of lives that will be saved, the therapies and medicines and technologies that will be developed, etc. Nothing about the concept of the "singularity" registers, just fear about their next paycheck - a fear to be expressed to their government at that, instead of killing progress that may one day save their own or their children's lives.
We volunteer you for AGI, and myself, and everyone else, because our selfish shortsighted nature and fear for our jobs are not worth the lives and suffering and death and disease of billions of people to come. It's sad that we lose our jobs - but it is the price we pay for the cure to everything. The best we can do is implement UBI while we make this transition as a species.
Trying to stop the development of AGI out of this fear is akin to being a plague doctor trying to stop the development of penicillin so as to not be out of work. I get it, we are all afraid for our jobs, but their importance does not stand a candle to the reduction in global human suffering AGI would be able to net us.
We are all literally going to die, you hear me? You, me, our children. It's not going to be pretty, because the lack of choice/consent around dying is the single biggest driver of human suffering, full stop.
But AGI could at least give us a choice here, a choice to spend many decades more in healthy life with those we cherish and love, fully provided for and safe and secure, and that's worth more than any job.
Saying there's a singularity in the future, when that's so very far from a guaranteed future in many people's eyes - who's to say we don't end up with a dystopian Terminator future rather than a utopian Star Trek future? Your time machine is just as good as theirs. We have three choices for our future. The billionaires, corporations, and the government. One of those three is where the singularity is going to come from. Which do you pick? Unfortunately, all three a flawed in different ways, so utopian Star Trek future of the singularity is by no means even guaranteed inn the first place.
You want people to trust that one of those three groups will just grant the rest of us living in the singularity. For all of human history, every revolution, it's been the same - there's been new jobs created. Just wait for it to be different this time. Avoiding the question of if we'll ever develop AGI because it's irrelevant, the question is when will machines be able to fix themselves, because that's when no new jobs or follow on jobs need be created.
If no new jobs end up being created, then what?
One time, it'll be different. I don't think we're looking at that quite just yet, but it doesn't see that far off in the future. Openinterpreter is already fixing problems for me on my computer without much intervention, robots doing the same for robots really doesn't seem impossibly far off.
I just pointed out that rushing into something that could make things better has historically involved an awful lot of suffering for people. An awful lot of things that were promoted as short term suffering for some people to make life better for everone (that remained) turned out to be long on the suffering and short on the "better".
So here's my thinking about an alternate plan:
What if we moved forward in a way that both advances the tech and also doesn't force suffering on people? This is possible.
There is no evidence that any of your promises, maybes and coulds are anything but words, and as such you are gambling other people's lives. The idea that "maybe we can help people in the future if we hurt them now" is merely a delusion without evidence, a lie those that hurt people on a whim tell themselves to justify and hide their true motives.
As for benefits, likely for everyone. People don't work just to make money, people make money since they work and produce for society right? Making money was never the end goal, and if we can avoid more work, that should be better for society as a whole.
In short, automation shifted the amount that needed to be done in certain industries. AI could possibly make shift so large that there won't be enough work for everyone to do.
I’m really not trying to argue for or against any new technology. But I don’t see why any new technological change necessarily brings about UBI given the absolutely drastic reduction in the amount of labor necessary for human survival we already achieved in the mid 1900’s. So my question is still, why did the old forms of automation not bring about a post scarcity utopia (or at least a UBI)?
I wasn't trying to say that people don't work for money fullstop, but rather that money isn't what people are trying to get ultimately. People want things, things that make people happy and alive. They get money in order to acquire those things.
So mass prostitution for women and mass gang violence for men is what the younger generation gets. To the benefit of the few and the elderly. It's already happening.
How will the clients pay if AI has already taken away the jobs?
So the result is that most women and a few men will have offspring, the rest of the men are exterminated from the genome. A future generation of people having a lot of cousins, with all the genetical diseases it brings.
Every new significant invention means massive genetical extermination of males and a new cycle of power struggles. A stable society with monogamy and equality among people is a rare breather.
things that currently would have to be done manually, but aren't feasible