Machine learning systems that we have today have become so powerful and are being introduced into everything from self-driving cars, to predictive policing, to assisting judges, to producing your news feed on Facebook on what you ought to see. And they have a lot of societal impacts. But they’re very difficult to audit.
Archive
I have a few examples of things that are starting to happen. And you can see that AI normally stands for artificial intelligence, but I’ve often concluded it stands for “artificial idiot.” And the reason is very simple. It turns out that these systems are good, but they’re good in kind of narrow ways.
I see a set of constraints facing us in the future, and they’re all going to be very expensive. First is funding retirements for the Baby Boom generation. Second is continuing increases in the costs of healthcare. The third is replacing decaying infrastructure. The fourth is adapting to climate change and repairing environmental damage. The fifth is developing new sources of energy. The sixth is what I see as in all likelihood continuing high military costs. The seventh is the costs of innovation.
The computer is being used for so many things that I claim that we have to consider the computer as part of our extended phenotype. It’s just a part of a thing that has evolved with us using memes.