It’s a rare and rewarding turn of events when the irrationality of our opponents is captured in a spot-on new term. So let’s savor the emergence of “doomer.” Naturally, I asked Claude just what “doomer” means.

A “doomer” is someone convinced that catastrophe is coming and probably can’t be stopped. It spread around 2018 as an internet meme: a cartoon of a gloomy young man in a black beanie, chain-smoking, hopeless about his own life and the world. The label then broadened into a tag for fatalistic pessimists on particular topics. A “climate doomer” thinks climate collapse is unavoidable. An “AI doomer” thinks advanced AI will likely wipe out or subjugate humanity. It’s usually a put-down, implying the gloom is more mood than argument. Some AI-risk people have adopted it for themselves, though. The opposite camp is sometimes called “boomers,” or more often “accelerationists.”

In today’s Wall Street Journal, “The Weekend Interview” is “Why Did Technologists Become Doomers?”

It’s not just the slant of the headline that’s good. The interviewee, Zack Korman, is astonishing–not just in being on the right side, not just in supplying important information about the risks of AI, but in actually showing that what caused the hysteria is a philosophy! And guess which philosophy: “effective altruism”!

The “effective altruism” brand has been pushed for a decade or two now. So let’s pause to evaluate the term. There are two things to bear in mind:

1. Effective altruism is still altruism.

2. Altruism is the morality of death.

On one level, “effective” and “altruism” are incompatible. “Effective” means bringing about the goal. Altruism means: surrendering all goals. But on a deeper level, there is one way a morality of self-abnegation can be effective: suicide.

Returning to the interview:

Mr. Korman traces the origins of AI doomerism to a 2014 book, “Superintelligence: Paths, Dangers, Strategies” by Oxford philosopher Nick Bostrom. Elon Musk praised the book at the time.

Bostrom’s profundity may be judged by his serious claim that we may be living in a simulation.

This is a revival of Descartes’ idea of the evil demon who force-feeds you false beliefs, such as that what you see is real. Another variant is Bertrand Russell’s idea that the world and you could have popped into existence 5 minutes ago but equipped with a whole set of (false) memories of a past history.

Back to “effective altruism.” It includes utilitarianism, “the greatest good for the greatest number” and runs wild with it. “The greatest number” is taken not just to mean everyone on earth, but everyone who will ever live.

Effective altruists add to utilitarianism the concept of “longtermism”–spelled with no hyphen–which Mr. Korman defines as the idea “that future lives matter the most, so to speak, because there’s so many more of them.”

He explains:

there’s eight billion people alive today, but over the next thousand years, we might hundreds of billions of people that come and go [i.e., not just because the population might grow but mainly because 1,000 years is many lifespans]. And if something bad were to happen to prevent those hundred billion people from coming into existence, we’ve effectively committed genocide.

Do you get the implication? You have to sacrifice not merely for the people around the globe who are alive; you have to sacrifice to bring into existence future people. And those who do not come into existence will hold you responsible for their never being born!

This is not idle theory. Mr. Korman knows these people, was at Oxford around the time Bostrom’s book became a thing, and has a career in cybersecurity. He says “effective altruism” is behind the doomer notions of slowing, restricting, or stopping technological progress:

Effective altruist organizations are “very committted to this concept of catastrophic risk and how do you prevent it. And one of the big areas of catastrophic risk would be AI kills everyone.” That Mr. Korman concludes is why “effective altruists as a community are extraordinarily doomer.”

Many of this ideology’s adherents have ties to large AI companies. Anthropic’s president, Daniela Amodei, sister of CEO Dario Amodei, is the wife of Holdem Karnofskyu, who co-founded two large effective altruist nonprofits. Anthropic received early funding from effective-altruism adherents including Sam Bankman-Fired. . . With so many effective altruists on the payroll, it’s little wonder that AI labs are filled with doomers.

There’s more, but I just want to pass along this little-known information about the vaunted Hugging Face break-in by a “swarm” of OpenAI’s agents who were being tested. Mr. Korman says the incident was due to a lack of “basic cybersecurity”:

OpenAI has said explicitly that if their monitoring was running, this would not have happened, but they weren’t running it.

I had heard it claimed that the Hugging Face incident was due to inadequate preventive measures, but I certainly did not know that OpenAI has stated that explicitly. Mr. Korman used to be a doomer himself, but learning more about the field–and the weird ideology behind the doomers–have converted him to be an AI booster. And he is rightly more than suspicious about the power that would be given to “regulators”:

Some, including Mr. Gates, have called for an international regulatory body to manage AI. “I think that’s quite a scary concept,” Mr. Korman says.

Correctly, he believes that legal liability law is the proper vehicle for handling such risks as there are– risks which are not that different, he says, from the risks posed by older technologies, as in the case of the automobile.