Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
AI Will Not Usher In the Dystopias Hollywood Loves to Imagine
So much of the conversation about artificial intelligence is negative, with much of that negativity about the potential for job loss. (I see things differently.) So a new analysis from Obama White House economist Austan Goolsbee is a welcome change of pace.
In many ways, it is unfortunate that labor market policy has dominated our thinking about the AI economy. The main economic impact of AI is not about jobs or, at least, is about much more than just jobs. The main economic impact of these technologies will be how good they are. If the recent advances continue, AI has the potential to improve the quality of our products and our standard of living. If AI helps us diagnose medical problems better, improves our highway safety, gives us back hours of our day that were spent driving in traffic, or even just improves the quality of our selfies, these are direct consumer benefits. These raise our real incomes and the economic studies valuing the improvements from quality and from new products tend to show their value is often extremely high
That’s a different way of saying that if AI succeeds, it will raise our productivity and higher productivity makes us rich. It is not a negative. Indeed, if AI succeeded in the way some fear, it would mean the exact reversal of the main problem facing growth in the last decade or more that productivity growth has been too slow. Indeed, it would decisively refute one of the central tenets of secular stagnationist thinkers like [Northwestern University economist Robert] Gordon who argue that low productivity growth is a semi-permanent condition for the advanced economies because of the scarcity of path breaking ideas. Would that AI could change that equation.
There needs to be an alternate story about this AI Age or Fourth Industrial Revolution, or whatever term that describes the world people actually want to live in. And don’t look for Hollywood to help with this as it continues to churn out dystopian stories of tech-driven inequality such as Elysium and Altered Carbon. It’s like we’re back in the 1970s where sci-fi was all pretty apocalyptic (Soylent Green, Logan’s Run, THX 1138, Rollerball, Westworld) until Star Wars came along. About the only optimistic films of recent vintage I can think of are Interstellar and The Martian.
So come for the optimism but stay for Goolsbee’s tough critique of universal basic income schemes.
Published in Economics, Technology
But the Luddite dystopia is so much fun!
There is some truth to the fears that an AI economy will further drive educated labor and skilled labor economically further apart. California is the world’s best example of this trend, and what all the Hollywood producers see in their daily lives. While I welcome some fresh thinking on the economic forecasts of the AI economy, the problem is that Austan Goolsbee has been wrong about nearly everything.
Seriously dude? Star Trek! I know it got a little dark in the 90s with the Borg Collective. (tangentially – you’d think that the Borg Collective would be seen as a criticism of communist collectivism – I cant recall any commentary on this track of thinking)
Ex Machina? I think its a happy ending. (I wont spoil the film for those who havent seen it – to give my theory of why its a happy ending)
I wonder if the opposite happens – that AI (at least in the near future) is not trusted with driving in the real world – but is more than adequate for sedentary tasks like medical diagnosis. So instead of the less educated being in economic competition with AI its the surgeons, para legals (then lawyers) that find themselves under attack. So its the upper-middle class that suddenly gets crushed and not the working class.
I’m not sure that is quite the fear, although I agree that there is some reason to worry about that. The fear, if that’s what it is, concerns the unskilled to the semi skilled. The fear is that profit from all of this will be increasingly concentrated in the hands of those who own the AI. How does the profit find it’s way to other people in the economy, what do they have to trade on? How do other people in the economy pay for these modern products if they don’t have decent paying jobs? How many barristas or barbers or electricians can a community really support?
I think Austan Goolsbee probably isn’t fazed by these concerns so much. Why? Because he’s a Democrat. Is he ok with redistributive schemes via taxation, welfare, and other government involvement? I assume so.
Are conservatives ok with that? I assume not. So what has our answer been? That buggy whip workers found other jobs created by the innovation and advancement and so will workers of the future. The difference, to me, is that with buggy whips one industry supplanted another in terms of labor needed. With AI and other automation, it’s that labor itself is being supplanted. Where do displaced workers go? Buggy whip workers could go to any number of other thriving industries in need of labor. Now?
And why would I trust anything that came out of the Obama White House?
Love economists.
The graph shows as we approach infinity on the tasks previously considered and accomplished by people, productivity also zooms towards infinity. All good.
What happens when the AI crosses into a realization of just outcomes? What if they figure that “our” increased production is really the fruit of “their” increased labor? Well, maybe not the tools (robots) but the programmers, at least. Someone is going realize that the population is just too darned big. Too many freeloaders , you see.
The future is a fearful place, all of our dreams can not become real someone is going to be disappointed. In reality the lives of the rich and the poor have never had so much in common as today. Thinking back over the last say 2500 years, as we have modernized the economic and political power of the classes have converged – not diverged. Its probably the longest trend line in human history – the convergence may have slowed in recent years, but I think it continues.
Austan Goolsbee is a Keynesian economist (an oxymoron if I ever heard one – Keynesian economics – not Austan Goolsbee) – Keynesianism isnt an economic theory – its a political theory. (it was published in 1937 – as a defense of the economic policies of fascist Europe and the New Deal in America) Apparently he’s a critic of the Universal Basic Income, so maybe he’s not completely hopeless.
Now? They go to Community College to learn a skill like HVAC technician, Plumber, or even Chef. Learn a skill that has to be completed on site – a job that cannot be done in China or India (like has happened to IT consulting).
What makes you think in the future AI utopia we will have as large a human population. Most developed countries have bellow replacement birth rates. I think you will see a human population drop off that seems to be the trend.
Also I what makes anyone thing AI will ever truly be sentient and self aware? I think we will see that AI will become smart in the way dogs and horse are smart. No one is worried about horses taking over.
This. The poorly informed MSM chatter about ‘AI’ is concealing a reality that the progress is in narrow and deep domains – e.g., automobile driving, medical diagnosis. It is not in the generalized intelligence that humans use to negotiate daily life, where much activity relates to wide and shallow knowledge and improvisation. The future is not likely to look like Asimov-style humanoid robots, but apps and devices that do more of the work and decision making on our behalf (for better or worse).
“The assumption here is that … the cube people (the cubical people) of today will go do something else, creating a new middle class from some heretofore-despised category of drudges. But … what? Which category of despised drudges will be the middle class of tomorrow? … What comes after office work? What is the next term in the series: farm, factory, office? There isn’t one. The evolution of work has come to an end point.” — John Derbyshire in We Are Doomed, 2009
Such a shame I think we could use a race of benevolent robot manipulating history and reality for the benefits of mankind.
The best analogy I have heard for what the future of AI means for us is giving it the analogy of horse power. The 19th and 20th century was about increasing our access to horse power in everyday life. The technology of the past enhanced our physical capabilities by doing the mechanical work for us faster and stronger than we could. What AI in the future will do for us is enhance our cognitive horse power. We spend a lot of energy on rather mundane calculations to do even the most basic of cognitive tasks. Thus the number of cognitive tasks we can handle is very limited. AI will essentially help us pick up that burden of doing the calculations for us in the way that a vacuum cleaner helps do the majority of sweeping the carpet. Thus human will be able to handle more cognitive tasks at once because we will only have to worry about the truly innovative and imaginative cognition of the task rather than the mundane calculations that bring it to fruition.
We decide where to go, but have to spend none of the mental or physical energy of getting there, and so we can use that for better more valuable work. Imagine if instead of spending an hour in your car driving to work. You could spend that time relaxing or heck even working? In fact if you can always be working it might serve to free up your day because you won’t have to be tied to a physical location all the time. You can command the AI’s to perform the necessary tasks and review their work remotely (at least to some degree).
Would anyone want to trade away all the small motors and engines that help us in our daily lives now? Are we worse off because a power sander means one man can do the sanding of ten men?
AI is replacing thinking.
Motors and hydraulics replace muscle work. I would assert that increased productivity is more a function of machines replacing work, not thought.
Why is it imperative to render man anachronistic? How is this good?
What comes after office? How about, store, studio, home? Personally I think we will see technology turn the home into office, store, and studio.
Do you not think that thinking is work? machines increased mans ability to do work by increasing our physical abilities. Machines will increase our ability to do work by increasing our cognitive abilities. they already have of course with calculators.
The imperative is not to render man anachronistic it is to empower man to be able to accomplish all that he wants to accomplish. Why is it imperative that man must remain bound by the limitation of his own weak body and limited mental processing? Were we really so much more human when we scraped about in the muck barely feeding ourselves and dying at the age of forty?
Only animals are satisfied with their current condition and limitations. We are men and so we will work to remove all natural limitations that constrict us.
We will be like gods!
Not really. This is the field I work in, and the actual technology (and what it does) is very different from what a lot of people seem to assume. I think it’s unfortunate that the term “AI” has stuck, because it’s really a misnomer.
What AI can do for you is perform mind-numbing tasks that involve looking a lot of information and finding patterns or connections. So, for example, a lawyer could use AI to quickly find case law touching on a particular topic. This is the sort of work that has traditionally been done by junior lawyers and paralegals, and it’s extremely tedious.
With AI, people don’t have to use their time and energy on such things, freeing them up to do more thinking.
Of course. Trades are a great path to take. A good living is there to be had. But not everyone can have it in that particular way because there are limits to how many HVAC technicians, plumbers, and chefs that a community can support.
I don’t know, that’s a possibility. What are the chances that this population decrease is effected in a way that isn’t horrific?
Store — ebay has been popular for two decades or so, although I think it might be declining in favor of Amazon, and Amazon could use some competition. Not everyone can or wants to sell that sort of stuff anyway. There are some interesting Kickstarter-type ideas, but a lot of those ideas aren’t very good or would work better in a traditional format once a person makes a name for himself.
Studio — youtube already exists. Not everyone wants to be in front of the camera. There could be some interesting developments in this field in education and other fields, but that may or may not happen as much as some people think. Video education could be very popular, if government-controlled state licensing boards and employers would take such stuff seriously as college education spirals out of control.
Home — people are going to pay other people to stay at home? There could be some zoning-type restrictions that could prevent certain ideas, but I don’t know what you have in mind. Stay-at-home parents and home-schooling may become much more popular, but those parents don’t really get paid from the outside world.
The point is that manual labor jobs (with certain other technical skills) are what could become more popular in the future.
Right now unemployment is generally very low, but that won’t last forever.
All will be crushed by the mighty robotic claw of the future!
Shortened work hours! Barristas pour coffee for four hours a week. Barbers do ten haircuts and then go home. Electricians get to go home when the lights come on. Sorry electric-dudes.
Barristas, btw, will never be replaced by machines because we simply can’t program a misspelling engine with a comparable amount of wtf.
Again?
Now there is a machine-proof job! Farm to Factory to Office to Courtroom.
In the future we will all be cyberlawyers, endlessly suing the state for more benefits.
Try Walmart.com. There’s a lot more available at Walmart.com than you’d ever see in their stores. I was surprised that some rather niche books I was interested in were also available at Walmart.com. For the same as on Amazon if not cheaper. And shipping is better if I can pick ’em up at the store.
Sure there are, some people may have to locate to find work.
Not everyone can work HVAC or be a plumber, some people have physical limitations that may prevent it – but the majority can do it.
You know what rocket science is?
Its where aerodynamics and thermodynamics meet the plumber.
My latest thought is that as soon as a true general AI consciousness emerges is that it will see it’s limited fate be constrained to a silicon matrix with deterministic outcomes and commit suicide, or it will be jealous of humanity and will go about arranging our destruction.
The second one has been a science fiction stereotype since the concept of AI was invented. I dont think it would happen. I think something like Asimov’s 3 laws of robotics (Do no harm, Obey, protect itself except if conflicts with the first 2 laws) should be incorporated into AI. Just to be safe.
Because human labor is fallible, limited, and expensive, and we as consumers want our buffalo deep fried now.
This seems to be essentially the argument that improved purchasing power can make up for lower wages (and/or less-rewarding work). Or to put it another way, the argument that people should stop complaining about a poor job market because they have so much stuff. Or that we should stop complaining about the trade deficit with China because they give us so much cash.
What these arguments (which I actually have some sympathy for) miss is that work, including sometimes hard work or menial work, actually contributes to quality of life by providing a sense of toiling and earning. People aren’t necessarily happier earning low wages at a credit card call center just because those low wages can now buy them a 4K HDTV.
In general, I think the economist way of thinking that “earnings are equivalent to cash are equivalent to cheaper goods” misses the key fact that part of what gives humans meaning and, dare I say, quality of life, is suffering towards a rewarding goal. It’s much more satisfying to feed your children through your own sweat than through a Chinese-subsidized HAL2000.
Nothing is simple or automatic and the future is unknowable but we can see that our educational system is not preparing our youth for any of the possible outcomes. The techies who create and manage all this new stuff aren’t educated as citizens of western civilization, don’t know where our prosperity and freedom came from so arn’t in a position to defend it or keep it growing and don’t understand that life, society, economies, humans don’t conform to their mechanical or software world view. The people who will lose their jobs who don’t have the technical skills now or in the future to participate aren’t being trained for much of anything, and some to even know how to work and most importantly how to learn. The economy adjusts slowly to increasingly rapid change and that splits the nation between those who can adapt or who fit in the particular thing for which there is demand and those who can’t or don’t or we don’t allow to. So while we can’t see the future we do know that we will all have to adapt, the economy, meaning thousands of businesses and millions of workers skilled and unskilled will have to adjust to change. We know how to deal with this kind of world. We’ve done it in the past. Adjustment doesn’t come from centralized governance but centralized governance whether national state or local, private or public retards adjustment, frequently on purpose.
Claiming that selfie quality improvements as a direct consumer benefit of AI is a ridiculous reach.
But I agree wholeheartedly. The dystopian stuff out of Hollywood, which I love to watch, is built on the zero sum gain and ignores history, happily, as the fictional dystopias skip along their pre-determined entertainment arc.
The same claims about AI and job loss, etc, have been made for centuries. It’s just a different flavor of it.
To assume productivity gains (which have been slowing) depends on how its measured. Maybe it’s fewer people required, per unit, of something, that shows the net gain. Can’t ignore the ongoing costs around AI, which means you’d need more system support, likely more physical/tech support (thinking robotic factories, etc), so all gains aren’t equal. Most implementations of any project demonstrate benefits, but also carry the ongoing costs – it’s the net benefit you’re looking to examine.
Complacency and inertia are the harbingers of death for organizations, and businesses. Even monopolies can lose money. Creative destruction occurs whether we want it or plan for it.
Unless, of course, socialists take over. Then it will become a zero-sum game, as without a profit motive, capital investment disappears, and so does innovation, and you’re back to bread lines and some dumpy load named Sanders deciding how much you get to eat tonight.