AI and the Duplicity of Silicon Valley
Technology billionaires and enthusiasts are breathless about AI. AI will herald the Singularity. AI poses a “risk of extinction” to the human species.
This is silly.
I’m not saying that AI will turn out the same as the steam-powered loom that had the Luddites up in arms in the early nineteenth century. It’s true that, throughout history, technological advances have usually ended up helping most people in the long run while disrupting the lives of many people in the short run. But I actually agree that this time is different.
The disruptions that AI will cause, however, are much more mundane than either a new generation of supercomputing humans or a takeover by robots. As Randall Munroe said, the immediate problem isn’t the enslavement of humans by robots, but what powerful humans choose to do with powerful robots.
The societal implications of robots that are better than humans at almost everything are not promising. Until recently, human society has been able to absorb the progress of technology because, each time a machine got better than humans at doing X, the humans figured out how to do Y, which was more valuable than X. So, for example, mechanized agriculture eliminated farming jobs. But mechanization (and the reduced need for farm labor) also made possible the twentieth-century factory, which created manufacturing jobs with higher productivity than the old farming jobs, meaning a higher standard of living for workers. Yes, there were always disruptions. Factory workers who lost their jobs in the 1990s to outsourcing could not magically become software developers. But, over the decades and in the aggregate, things got better for most people.
In the future, I fear that the set of jobs that humans can do better than AI-powered robots will be too small to provide enough work for everyone, or will be too low-skilled to provide a decent income. (Machines are pretty bad at doing things like making beds in luxury hotels, for example, or bringing $100 entrees to restaurant tables.) The problem is that we don’t have a social organization equipped to handle this situation. The value created by the robots will accrue to the very few people who design, program, and own those robots. Everyone else will either be out of work or will be in the service industry catering to a shrinking number of increasingly rich people (the robot designers, programmers, and owners) who can afford those services. This has already happened, as the workforce has transitioned away from factory workers to hotel housekeepers, restaurant kitchen staff, sales clerks, and home health aides. AI will only accelerate the process.
Things don’t have to be this way. We could have a society where the enormous surpluses that could be generated by AI are widely shared among everybody, including the people who are displaced by robots — a society in which the people who design and program the robots are rich, but not so extravagantly wealthy and powerful that the rest of humanity has to compete to entertain their demand for luxury.
I’ve written about such a society and what it will take to achieve it. But technology titans like Sam Altman — a prepper whose solution to major threats is to look out for number one — have no interest in using their power to make our society better. Instead they use the specter of Skynet to prevent people from seeing what is really going on.
