Opinion: Konrad Körding and Ioana Marinescu from the University of Pennsylvania argue artificial intelligence will likely have a limited impact on jobs because of the realities of physical work
The singularity narrative posits that intelligence can design robots (and robot manufacturing processes) that quickly bootstrap to arbitrarily large influence on the physical world. The model presented here seems to ignore that possibility for unclear reasons.
This pushback, in my reading, assumes that (a) current machine designs are bad and could be much less resource intensive (b) making new designs does not require machines (e.g. for testing). (c) the bottleneck is not the machines to make new machines. I hear your argument but I think I don't see evidence.
Machine designs of 30 years ago were abysmal by today's standards and most of the story for how we improved them comes down to relieving intelligence bottlenecks: better hypotheses, analysis, and yes some testing (where test design is a major determinant).
So yes, the singularity narrative assumes there is a huge amount of invention/innovation left to do, but assuming we're at the end of the tech tree or especially bottlenecked by the availability of human hands for testing seems much less plausible given recent history.
Imagine you're trying to bat in baseball. You have ultra-high resolution optics, think about 1000x faster than any human, and control a humanoid robot body with high-precision mechanics. Are you really telling me that this wouldn't increase the chances of a hit? It wouldn't go up to 100% for sure, there are physical limits. But they are way beyond current human skills. So at one point, humans will be completely outperformed by smart robots in Baseball like we already are in Chess and Go.
Even if some physical jobs cannot be automated so easily, there are millions of people doing things like truck-driving that are already being automated. The industrial revolution has automated physical labor first, now AI comes for the white-collar jobs and humanoid robots come for the more complex physical labor. I very strongly doubt that all those people will find new physical jobs, let alone jobs they'd like to do.
All this doesn't necessarily have to be bad if we can manage the transition and redistribute the wealth so nobody has to work anymore and we can all do nice things together (provided the AI will remain docile and not take over control completely). But I don't see that we're on the way to that future.
This gives me no hope: "Think of AI optimizing marketing and food purchases for restaurants". Remember all the people who would be involved in that marketing - copywriters, designers, photographers, models potentially. They've all lost their life's work. And now they're going to dig ditches? Fuck all things AI.
The singularity narrative posits that intelligence can design robots (and robot manufacturing processes) that quickly bootstrap to arbitrarily large influence on the physical world. The model presented here seems to ignore that possibility for unclear reasons.
This pushback, in my reading, assumes that (a) current machine designs are bad and could be much less resource intensive (b) making new designs does not require machines (e.g. for testing). (c) the bottleneck is not the machines to make new machines. I hear your argument but I think I don't see evidence.
Machine designs of 30 years ago were abysmal by today's standards and most of the story for how we improved them comes down to relieving intelligence bottlenecks: better hypotheses, analysis, and yes some testing (where test design is a major determinant).
So yes, the singularity narrative assumes there is a huge amount of invention/innovation left to do, but assuming we're at the end of the tech tree or especially bottlenecked by the availability of human hands for testing seems much less plausible given recent history.
Imagine you're trying to bat in baseball. You have ultra-high resolution optics, think about 1000x faster than any human, and control a humanoid robot body with high-precision mechanics. Are you really telling me that this wouldn't increase the chances of a hit? It wouldn't go up to 100% for sure, there are physical limits. But they are way beyond current human skills. So at one point, humans will be completely outperformed by smart robots in Baseball like we already are in Chess and Go.
Even if some physical jobs cannot be automated so easily, there are millions of people doing things like truck-driving that are already being automated. The industrial revolution has automated physical labor first, now AI comes for the white-collar jobs and humanoid robots come for the more complex physical labor. I very strongly doubt that all those people will find new physical jobs, let alone jobs they'd like to do.
All this doesn't necessarily have to be bad if we can manage the transition and redistribute the wealth so nobody has to work anymore and we can all do nice things together (provided the AI will remain docile and not take over control completely). But I don't see that we're on the way to that future.
This gives me no hope: "Think of AI optimizing marketing and food purchases for restaurants". Remember all the people who would be involved in that marketing - copywriters, designers, photographers, models potentially. They've all lost their life's work. And now they're going to dig ditches? Fuck all things AI.