I wanted to make one remark about Noah Smith’s recent observation on AI’s “sales pitch”:
“Hi. Do you have a moment? I’m from the Cursed Microwave company. Our product is much better than a traditional microwave. Not only can it automatically and perfectly cook all your food, it also microwaves your whole body, so you and your family are paralyzed and unable to ever work again. Don’t worry, though, because when everyone has a Cursed Microwave, our society will probably implement Universal Basic Income, and you and your children can just go on welfare! Oh, by the way, we estimate that there’s a 2 to 25 percent chance that our microwaves will put out so much radiation that they destroy the entire human race.”
If a door-to-door salesman gave me this pitch, I would gently see him out the door, and then quickly call the FBI.
Quite to the contrary, it’s a great sales pitch if you define “great” as one that converts paying customers. But neither Mr. Smith nor ordinary people are the intended audience.
It’s a compelling pitch to short-horizon, competition-constrained decision-makers even if the long-term equilibrium is unclear or unstable. Phrasing it more kindly the pitch is:
Using AI you can accomplish the same results you are now with fewer employees. We can improve your bottom line.
It’s perfect when you identify its audience correctly.
Assuming there are mass layoffs, it will plunge the country into a Greater Depression. Since people could not repay loans, it would collapse the financial industry. Many AI run companies would cease to exist.
As to AI workers, how long before the board of directors realize they can hire AI executives? Why pay Elon Musk a trillion dollars when AI can do the same thing?
While the AI workers can replace humans, can they produce the same or better products? If not, will those few people with jobs still buy those inferior products? Could some newly unemployed worker start a new business with AI workers?
Will AI workers replace the humans to make improvements in AI code? If AI is as ubiquitous as predicted, will the price drop?
With the financial industry collapse the stock market will crash. CEOs whose income was based upon stock and stock options will not fare much better than their employees.
Anything that is rule based can be replaced, and that is a good thing. Trying to replace anything else will result in eventual failure. AI must operate within a framework. No matter how autonomous it may seem, it operates according to rules, and humans excel at finding glitches and exploits in rules.
Demonstrations of robots that can walk, jump, and climb stairs are a stage show. They are only able to perform under specific circumstances. Roll a ball in their path, and see what happens. AI is no different.
I may be naive, but i do not see an AI apocalypse.
Even more to the point given the sharp rise in executive compensation the value of replacing managers with AI is that much greater. The barrier is that decisionmakers are very unlikely to decide to replace themselves. I believe that’s true in most of the areas in which the potential of LLM AI is high.
You wrote:
I agree. I see a management-produced apocalypse that will be labeled an “AI apocalypse”. Who will stop it?
@Dave Schuler
… Who will stop it?
It is like a frog being boiled, slowly. When it becomes too uncomfortable, it will jump out.
I am not familiar with the AI industry, and some of this may be occuring already.
At some point, companies selling AI code will soon begin licensing it on a per worker basis, and too increase revenue, they will break down the code into smaller pieces to make more “workers”. AI agents will do fewer and more specialized tasks. Eventually, the subscription model will be applied to AI agents.
Another issue: will AI workers be required to follow the law. Will replacing humans with AI accountants eliminate tax fraud. Will AI self-driving cars be required to follow traffic laws.
A lot of human activity revolves around glitches, exploits, and cheats. Will AI be given the same freedom? If not, will things be better or worse.
Eventually, there will be knock-off and/or imported AI agents. Why would management hire American AI workers? Management has no problem with outsourcing to China or hiring illegal aliens. Why not use pirated Mexican code or have the work outsourced to China, India, etc.
I have a lot more, but it is coming too fast for me to write. Anyway, there are a lot of issues that have not been and cannot be anticipated.