5,847 karma · joined May 24, 2019
Usually the first additional link to an article is an archive/non-paywall link. That the two links went to the same thing was really the point of my question.
See citizens and ICE.
Public works projects need funding and purpose. I'm not optimistic about what purposes the make-work funding gets applied to these days. Public beauty, cleanliness, health are not in the priorities.
The human system relies on us doing stuff for each other, mediated by "money". We get "paid" by some other human for doing stuff for them. "Stealing" is discouraged. "Being given" is rare. "Money" originates from fiat, goes where it goes.
Humans unable to do for others don't get "paid", get starved/sick etc, experience pain, cause trouble and disappear eventually. When doing stuff for people is not done by people, the pay cycle breaks. When machines do a substantial fraction of economic activity (intangibles like decision-making, writing, answering the phone, etc. as well as transport), there is no need for very many people. Absent the need, there will be no support. The farms won't give their food away, they will arm themselves against the hungry.
There will be some human work. So long as machines cannot equal human touch and experience combined, there will be medicine. There will always be sex work. But who affords it?
Possibly racial disparities will diminish. A machine might actually give equal medical care to, or equally arrest, the Black and white person. Or not, depends on the training. Look who's developing the AIs.
Economists thought that economic development means democratic development. They were wrong, see China. They thought that economic liberalization means good governance. Wrong, see Russia. Economists are right about things until they are wrong, and we can all run off the cliff with them.
When the population drops (war, starvation, refusal to bear children) to the level needed to run the machines, the system will stabilize. Pretty much the horse scenario.
Comsider the tracking potential.
Now consider that they can perform textual analysis at a scale large enough to figure out who you are based on your linguistic fingerprints across HN, Meta, Github, forms you've filled out for you child's school, your doctor, or wherever else a service provider has used them.
These are "AI" companies but monitoring is what the AI is and will be used for.