William Miller, The Great Disappointment, and Artificial Intelligence Systems

This past week there were numerous articles from and interviews with various AI bros saying that within 12 to 18 months, AI will replace white-collar work and humanity must adjust.

When I read this articles, I wasn’t reminded of the Singularity, of AI, of Skynet, or anything technological.

Instead, I thought of a preacher named William Miller who died a about a hundred and ninety years ago.

William Miller came out of the Second Great Awakening, which was one of the waves of religious vitality and furor that grip America every so often. Miller almost died in combat as an officer in the War of 1812 and saw one of his men killed in front of him, which understandably left a lasting impression. His experiences led him to an examination of mortality that resulted in a fervent Baptist conversion. He also became convinced that he could calculate the date of Christ’s return from the Bible, and decided that Christ would return on October 22nd, 1844. By then he had a substantial following, and on the day his followers gathered in their churches to await the End of Days, many of them having already given away their possessions.

But nothing happened. Miller’s movement collapsed and most of his followers abandoned their beliefs, though some splinter groups eventually evolved the “Adventist” branch of American Protestantism, of which the Seventh Day Adventists are the most prominent.

Now, when Miller is discussed online, the usually tone is to laugh at the religious rubes from the benighted past, so unlike us enlightened and savvy moderns. But I think the truth is that Miller succumbed to a universal human impulse. Every generation thinks that it is going to be the last generation, or the generation that will see the culmination of history, whether religious or secular.

For example, when I was in my early twenties I knew a very religious woman my own age who was convinced that the world had become so wicked that it would end by the time she was thirty. A few years later I met another woman who thought global warming would ensure the collapse of the ecosystem and the end of the food chain by the time we were thirty.

However! I have not been thirty for a rather long span of time now, and for better and for worse the world grinds on.

Nor is this an impulse limited to my own generation. People who came of age during the Cold War thought the world would end in nuclear fire during their lifetimes, and a little after that from global cooling. Lesser examples could be seen in the Y2K scare. Throughout the Middle Ages and the early modern period, it was common for peasant revolts to be led by charismatic preachers who predicted that soon all thrones would be overthrown and Christ would return to judge the living and the dead.

Because of all these examples, I am certain there is a universal human impulse to believe that the world will end in our lifetimes. I think this comes from partly a combination of fear and hope – fear of the future and the end of the world, and hope that one’s life will be lifted out of the mundane in the final fulfillment of history. You don’t have to get up and go to school or work tomorrow if the world ends!

But the truth is that the world most likely is not going to end and you and I are probably going to have to get up and go to work tomorrow.

I think the hyperbole about AI comes from that same impulse – this idea that one is living to see (and participating in) the apotheosis of history, when what one is in fact doing is using a money-losing chatbot that frequently gets things wrong.

To be clear, AI isn’t going to wipe out white-collar work and it isn’t going to cause the collapse of society, though like cryptocurrency it will cause a lot of harm without much benefit. AI simply isn’t good enough and doesn’t do what its boosters say that it can do. There are numerous people who, in my opinion, are accurately explaining and pointing out the many flaws in AI and in the economic bubble. Just as there were people who predicted the fall of the Soviet Union, the dotcom bubble, the housing bubble, the criminal activities of FTX, and the flaws of cryptocurrency, and were frequently derided as cranks until subsequent events proved them right.

So why all the hyperbole around AI? I think part of it is the end-of-days impulse mentioned above.

The rest of it, I’m afraid, is sheer crass desire for money and power. Why are all these tech companies burning unfathomable sums of money on AI when it is obvious, painfully obvious, that the bubble is heading for a crash? After the dotcom crash of the early 2000s, the Internet companies that survived eventually evolved into the tech titans of our day – Amazon & Google come to mind. All these different AI companies and boosters are hoping that their company is the one that survives and becomes the next titan conglomerate of the 2030s.

Admittedly, I think that is unlikely. I think the most probable outcome for the current model of AI (LLMs and generative AI) is that it ends up like cryptocurrency. For a while crypto advocates thought it would overthrow central banking and led to unprecedented freedom and prosperity. However, while there are many valid criticisms to be made of central banking and fiat currency, one of their advantages is that they do a good job of shutting down a lot of the scams that crypto easily facilitates. For all the glowing promises of its boosters, the primary use case for cryptocurrency has been to cause economic disruptions and to facilitate crimes and scams.

I suspect AI will probably degenerate down to a similar state once the bubble pops. It can’t do all the miraculous things that its backers promises, the money is going to run out eventually (and it will inflict a lot of economic damage on the way out), and like crypto, AI will mostly have negative uses. Likely its most common use cases will be to help students cheat on exams, make stupid political memes where someone’s least favorite politician (whoever that is) is shaking hands with Emperor Palpatine or Thanos or whoever, engage in mass copyright infringement, and to scam seniors out of their savings.

So if you are disturbed by the rhetoric around AI, take heart. When you read an article from someone announcing the glories of AI, remember they are most likely 1.) seeking money or power, or 2.) are like William Miller’s followers the day before October 22nd, 1844.

-JM