Since the introduction of Chat-GPT a few years ago, it’s seemed like AI was unstoppable. As tech-bro’s exclaimed the virtue of the technology and stock markets soared via historic splurges in AI investments, we seemed to be tumbling into a world were “smart” machines are poised to invade every corner of our lives.

Over just the last few months, however, something amazing is happening. People are making it clear that replacing everyone’s job with AI’s that might also lead to the end of humanity might not be such a great idea. Since I am watching this with as much surprise as everyone else, I wanted to take a beat and think about what is happening.

To begin, I think AI technologies are amazing. I have been using them in my physics research in ways that are startling and transformative for me. In particular, I have been using Large Language Models to help me gain a deeper understanding of features of General Relativity that had not been possible for me before (even though I have, literally, taught the class). The second place to start is that AI means many things. This is a point Princeton-based scientist Molly Crockett made at our recent Varela Symposium at the Upaya Zen Center. The Chat-bots many of us encounter are just one kind of AI technology.

Neither of these points, however, speaks to how these technologies will be deployed in society. That is where the real danger arises. I think the hype about “Super-intelligence” is just that. The problem is not that these are machines are smart like us or smarter than us. Instead, they are dumber. They are really good some things like pattern recognition. But they are also terrible at other things humans beings do which are essential for a functioning society (like feeling empathy). So the real danger is that these “less-than-us” machines will be jammed into every corner of our lives. Then we will be forced to meet them at their reduced level.

Of course the real problem is who is doing the jamming.

The most dangerous part of AI comes from the handful of ultra-wealthy, ultra-powerful people who stand to become vastly more ultra-wealthy and ultra-powerful if their plans for AI succeed. Over the last few years we’ve been told over and over again that the rise of AI in society is “inevitable”. Yes, all those jobs will just have to be lost but - in some miraculous way - everything will eventually be better. Yes, these machines might end up switching off your local power plant by mistake but, eventually, they will usher in an era of utopian peace and prosperity (and ignore their use in the terrifying prospect of autonomous weapons). Yes, these machines will require building zillions of data centers that will make climate change worse but don’t worry they will - miraculously - solve climate change too.

So, what’s fascinating to watch now is the beginning of people simply saying “No”. I have been arguing for a while that nothing is inevtiable in terms of how any technology gets used in society. That is always up for grabs. Genetic engineering, for example, had pretty strict guidelines put on it from the get-go and you still can’t clone human beings.

The rise of movements against data centers is really just a proxy for citizens beginning to claim the right to have a say in how AI appears in our collective life. I am in full support of these movements. There are ways we can still allow AI technology to develop without rushing to give its dangerous and untested aspects control over our lives and our societies. And I absolutely acknowledge that every technology has moral costs which we all try to navigate I have had my work vacuumed up by Anthropic without my consent and yet I used AI to make the image in this post. But AI tech is still new and there are many ways for it to evolve. Most importantly, the issue is the way AI gets broadly pushed into society by just a few people. There is too much at stake to leave these decisions in the hands of a few tech-bros.

That is why what’s happening this summer is both surprising and hopeful.

=====================

P.S. I am really interested in your thoughts for or against Please leave a comment on the website for this post or email me at [email protected]

P.P.S. I could not get this proof-read so please excuse typos etc.