Anthropic CEO Dario Amodei warned that within the next five years, AI could eradicate half of all entry-level jobs and increase unemployment by up to 20%. University of Louisville professor and computer scientist Roman Yampolskiy makes a darker prediction, claiming that 99% of jobs will be eliminated by robots and AI in the next five years.
These appear to be scary times indeed – AI is now creating images of people with the correct number of fingers and the future that was promised where AI takes our jobs is no longer some ominous event in the distance. Computer nerds just couldn’t help themselves.
We’ll start by noting that the ludicrous prediction by Professor Yampolskiy is wildly implausible and perhaps even stated in bad faith.
Robotics is a ways away from the sophistication needed to complete all of those jobs and even companies implementing AI for tasks that were predicted to be easy to replace are encountering integration problems – some even having to rehire humans after jumping the gun on AI. A recent MIT study found that 95% of AI implementations at companies failed. Nevertheless, it seems reasonable to predict that eventually all work will be replaced by robots and AI.
The most optimistic among us reply that inventions like the internet and past machines all produced concerns about their effects on the demand for human labor. As it turns out, rather than wiping out the need for human labor, these inventions simply allowed human workers to become more productive.
These optimists might point to these examples to argue that AI will have a similar impact – creating different jobs and increasing the productivity of human labor instead of replacing it. The clear difference though is that combine harvesters need a driver, lathes need operators, and the internet is used by workers. AI on the other hand, if advanced enough, is an almost complete replacement, not an enhancer.
We’ll set aside the possibility that AI somehow destroys civilization by more direct and explosive means and focus on what happens if that catastrophe is avoided. There are short and long term consequences that are distinct given our readiness. It’s highly likely that, in the short term, our lack of preparation with policies that address a modest (but still significant) jobs loss will result in high unemployment with little remediation leading to sociopolitical instability. I’ll address the short-term outlook in the next column.
But what does a future look like where we’ve arrived at near complete automation? Natural questions arise about what sort of economic system would survive if there is no need for humans and individuals aren’t able to sell their labor.
Of course, corporations have revenue in the first place because people have jobs and can pay for their goods and services. It seems self-defeating that corporations would eliminate so many jobs that people are simply unable to purchase their products.
As I see it, there are a few likely possibilities. The first is that a large tax is levied on corporations and the funds are distributed as universal basic income. Among the challenges here would be in determining what the appropriate standard of living should be for people.
The oligarchs and politicians would be wise to provide a life that is just pleasant enough to suppress our urge to guillotine them. If automation greatly reduces scarcity and we are able to tax corporations sufficiently, we can look forward to a future of idle time filled with pottery classes and bird-watching.
The second possibility is that corporations pivot to providing high-end goods and services – in this case, the oligarchs and shareholders would sell to each other, either leaving everyone else behind or providing a life barely worth living. Humans, being the resilient and intelligent creatures that they are, wouldn’t lay down and starve to death. In all likelihood, if people are forced to live in squalor as they look upon the shining skyscrapers from their slums, either a revolt would occur or an alternative market would form independent of the AI-using corporations.
Inevitably, because of competition and market pressures, AI will slowly seep back into either of these outcomes and humanity will be given another chance to organize itself in a way where everyone is given a good life. I like to think that we will eventually make the right choice.
Rafael Perez is a columnist for the Southern California News Group. He is a doctoral candidate in philosophy at the University of Rochester. You can reach him at rafaelperezocregister@gmail.com.