AI Will Rapidly Transform Labor, Exacerbating Inequality, Insecurity, Poverty
scottsantens.com
scottsantens.com
The cost of government programs is often tied to the cost of creating a useful administrative structure. When real humans are involved, they devise fiefdoms, and cost tends to climb. When we try to privatize them, it mostly shuffles around the problem.
When you say, "now the AI will do it", there's no specific limitation stopping "it" being the entire planning stage, down to step-by-step instructions for the team on the ground.
So when we speak of UBI, welfare, etc. - that's already outdated 20th century stuff. Think bigger. Every citizen gets a personal AI agent attending to their material needs, helping them navigate the laws, know their rights and request assistance. Why wouldn't they? That means that the marketplace - for jobs, consumer goods, etc. - becomes more of an abstraction layer interfaced with by agents. They don't need any specific human interface like a job board or a form to request unemployment benefits.
So, the picture of supply and demand rapidly changes away from the blunt instruments of money/wage/tax/investment towards scalpel-like innovations, complex deals that humans could never have negotiated. Among those will be policy programs for issues of poverty and retraining.
The author is naively mistaken in what dictates outcomes. Incentives dictate outcomes, there's a lot of flawed thinking and wrong conclusions made by the author.
Much of the argument relies heavily on survivorship bias and painting implausible examples. It also completely neglects any basic understanding of what happens with economic systems. Most specifically, how UBI would solve the economic calculation problem (point in fact it doesn't, it makes it worse because there's no price discovery). So what the author actually is promoting is a rapid approach to all the consequences found in the economic calculation problem, though they fall just short of saying so.
This reads more like soft propaganda than it does anything educated or rational.
This eloquently sums up a lot of discourse on AI.
There was a very good video someone posted not too long ago that covers what should be covered as an overview.
Link: https://www.youtube.com/watch?v=xoVJKj8lcNQ&ab_channel=Cente...
Any engineer knows as soon as the rate of change exceeds the reaction time, you can't keep up and failures happen. In the case of systems, the larger the system the bigger and more damaging the failure. We're dealing with societal mechanics, so failures may have global impact.
In a lot of western countries, there's already a tax system, from which welfare is distributed. It's arguable whether it is sufficient, but it is currently a good enough system.
UBI, in my opinion, fails because it disincentivizes the desperate from becoming innovative. It makes it so that there's zero risk in consuming more than you produce. The only circumstance in which this is allowable must only be when there's absolutely zero scarcity anywhere (e.g., humanity achieves near-infinite, free energy). We are nowhere near that.
People who are desperate are generally quite incentivized. They want to get out of their current situation. Being innovative does not require someone to be desperate.
Only circumstance? How about when people are looking to live a basic human existence but are unable to find a job?
Desperate people get myopic and risk averse. Stress focuses you on the next step at the expense of the long view. If escaping your pattern requires a longer term plan you are stuck. You might not even see the loop you are in anyway.
Isn’t that literally the poverty trap?
UBI might become low enough to be practical with a mix of changing attitudes around living “the bare minimum” and motivation to fix absurd expenses when possible.
Living in a dorm room in a rust belt town with a cafeteria and 1940s-level clinic might look less dystopian when it’s the totally free option.
Get that level low enough and you can expect most people can actually succeed in creating about that much in value because it’s absurdly low and any person would struggle to create a hard 0.
Even without UBI, if it were legal, I think an ultra cheap lifestyle option would be a popular niche. Machines can easily support a 19th century living standard for almost nothing.
There are plausible positions against UBI[0], but this is not one of them.
The entire history of innovation has come from people with enough resources and time to freely sit and contemplate and tinker and build elaborate contrivances for fun and curiosity. Desperation is not a valuable factor. The extreme contrary is so blatantly clearly true that the authenticity of that train of thought beggars belief.
Someone starving to death in a gutter desperate for shelter and a meal didn't invent calculus, the printing press, electricity, cars, computers, telephones, textiles, airplanes, rockets, antibiotics, vaccines, anesthesia, mechanical refrigeration, cameras, soap, pre-sliced bread, salt iodization, next day delivery, dynamite, microwave ovens, organ transplants, weather balloons, satellites, lasers, submarines, and factory canned beans.
Those things were developed by people who had comfortable lives with room for curiosity and exploration and a clear financial runway. Comfort is demonstrably more relevant to invention than desperation will ever be.
[0] - One could argue, for instance, that it will be very difficult for UBI to not just make rent increases in desirable areas soak up the entirety of the surrounding outlay without a slew of other policies, as property owners would just see more milkshake available to suck up through their very long straws. (A common counterargument is that people would suddenly be free to move to undesirable areas with lower cost of living, which is true but ignores the whole "desirable" vs "undesirable" part.)
> disincentivizes the desperate from becoming innovative.
I thought it was the opposite? If you look at the human history, most of the progresses come from the ppl who can afford to do try something else.
Luddites must be rejected because they're simply never right. Extreme poverty has never been lower in history thanks to automation driving the price of food to the raw cost of production.
Remember, 90% of the population used to work in agriculture toiling away. Now it's ~2% and yet poverty has absolutely plummeted.
A significant portion of the population today is toiling away in manual jobs, tedious jobs and pointless jobs. This will all be automated away leaving those people free to pursue their most human pursuits.
I work in software, and I cannot wait until AI does 90-99% of my job. I can't wait until I rarely write software manually. To me, writing software is the most boring part of my job. The best part is outcomes, and I can achieve 100x better outcomes if all the grunt work is taken care of by AI.
Luddites have never been right, and they've consistently been proven dead wrong. This is no exception.
> As someone who has been warning about the impacts of automation for a decade now, and trying to get people to understand that automation has already been impacting us for decades, I can tell you that this nuanced conversation basically gets boiled down to "Technology is going to end all jobs" versus "Everything will be fine." Both of those are wrong. The truth as usual, exists somewhere in the middle, and comes with a lot of "it depends" asterisks.
Address the content of the article, and argue whether this basic premise is true or not. Conversations about "luddites" have been had to death and are basically worthless.
> This will all be automated away leaving those people free to pursue their most human pursuits.
Citation needed. This is all _very arguable_ and incredibly arrogant to assert over and over that things are going to be fine because in the past it was fine (for flexible definitions of fine).
Claiming new automation and new technologies will affect poverty negatively is an extraordinary claim.
Everyone can agree that AI will dramatically lower food prices until they near the price of raw production, therefore it's pretty laughable to claim that the lowest food prices ever will coincide with more poverty.
It's a fundamentally Kaczynski-esque argument, that AI is evil and must be curtailed. Moreover, it's an argument that humans should be enslaved to agri fields, enslaved to cubicles and enslaved to dangerous oil rigs forever. Humans don't want to destroy their body toiling forever, they want to live.
Please provide some sources supporting the existence of this mythical treasure, this economic prediction on which “everyone can agree”.
https://locusmag.com/2022/01/cory-doctorow-science-fiction-i...
> In truth, the Luddites’ cause wasn’t the destruction of technology – no more than the Boston Tea Party’s cause was the elimination of tea, or Al Qaeda’s cause was the end of civilian aviation. Smashing looms and stocking frames was the Luddites’ tactic, not their goal.
> What were they fighting about? The social relations governing the use of the new machines. These new machines could have allowed the existing workforce to produce far more cloth, in far fewer hours, at a much lower price, while still paying these workers well (the lower per-unit cost of finished cloth would be offset by the higher sales volume, and that volume could be produced in fewer hours).
> Instead, the owners of the factories – whose fortunes had been built on the labor of textile workers – chose to employ fewer workers, working the same long hours as before, at a lower rate than before, and pocketed the substantial savings.
I think this is an important conversation to be having around AI. Deeper in the thread someone argues that AI will make farming cheaper and then drive down food costs. But, food isn't expensive because it's expensive to produce. Food is cheap and easy to produce these days. Food is expensive because Monsanto copyrights seeds, because John Deere won't let you fix your own tractor, and because the usa government pays subsidies to keep production down to avoid the market being flooded and crop costs plummeting.
The advancement in farming technologies didn't lead to the conclusion some of us find quite natural: one more thing we don't really have to pay for, like water or roads to walk on. Something so cheap to produce and distribute that our taxes can just take care of it, and we can always buy fancier versions of if we want.
For me I think this will be the effect industry to industry that AI knocks on the door of. The technology itself isn't the important change, it's how those improvements trigger redistribution of wealth, if at all.
I'm working on a blog post right now about how this might affect propaganda, when AI is turned to that purpose, and the risk therein of very effective propaganda written that convinces people that it's good that AI is being applied in a way that makes their lives objectively worse (as it drives down the value of their labor to almost nothing).