I was at a get together with a bunch of family members. And, as is the case with gatherings of this sort, we talked about a variety of subjects of almost a ridiculous range; Joe Pesci’s guitar playing in the shower, the theory that Alzheimer’s disease is an autoimmune disease, a recent positive experimental result in the search for dark matter, the likelihood that artificial intelligence will achieve self-awareness and a bunch of other stuff. But the topic that had the most separate conversations over the course of the day was AI and the data centers that are being built to support it.
And I heard a range of opinions. Everything from AI was the granddaddy of all bubbles that would shred the current tech boom once and for all, to the opposite opinion; that AI was about to create heaven on Earth where the productivity created by AI would allow us to live our best lives while problems like limited resources and affordability would evaporate in a world awash in material wealth. Well, who’s right?
I’m going to guess that both these scenarios are fringe ideas that don’t realistically represent what’s currently going on with AI.
Whereas there probably is a case to be made that the trillions of dollars of infrastructure being assembled to allow AI to “transform every part of human industry” may fail to provide a positive return on investment and disappoint Wall Street. But I would assume that would cause a bear market, not a complete meltdown of the world economy.
And as for AI creating an earthly paradise, I think it’s more likely that AI is going to cause a major disruption in employment markets that will be a net negative impact on everyday people until the changes caused by AI begin to equilibrate in the economy. At that point there will be a new reality where the higher productivity will allow for the standard of living to improve.
So, AI will be neither a panacea or a catastrophe. It will be just another change in the environment that human intelligence brings about as it attempts to increase its control over the physical universe and optimize the way that humans interact in what we call civilization.
But one thing I am sure of. We will all have to get used to talking to bots almost exclusively. And for this reason, I really hope that AI gets much, much better at training them to understand English. Every time I have to struggle through trying to explain a customer service problem with a machine that is as dumb as a bag of hammers I can’t help but think that the Unabomber had something in that manifesto of his.
But despite all this I guess I’m basically optimistic about the future of AI technology. I can already notice a great improvement in the quality of the answers I get from AI’s like Grok today as compared to a year ago. They definitely have made these things smarter and more reliable than before.
Oh, and on a personal note, I had one of my cousins buy a paperback copy of my book when he found out about it. I hope he likes it. He might tell his family. And he has a big family.