I asked ChatGPT to create an image based on chapter 10 of my 2014 novel Fugue in Ursa Major. Jake is in Washington, and all hell is breaking loose on a global scale. Jake is about to head for the hills. But much has changed since 2014. Click here for high-resolution version.
This morning there were yet more stories about how AI could exterminate the human race:
Anthropic researcher Jacob Coxon has quit the AI company, saying those building the technology “earnestly believe that it could kill us all by the end of the decade.”
If the people issuing these warnings were anti-technology outsiders, we’d do well to take them less seriously. But these warnings keep coming from people who actually are involved in AI development. There rarely are any specifics on how AI would actually do such a thing. A bio-engineered pandemic is sometimes presented as the easiest way. Another type of threat has to do with AI getting control not only of weapons systems but also life support systems such as water systems and the power grid.
We might take some comfort in the RAND Corporation’s findings that exterminating all human beings would be very difficult, because humans are so adaptable and there are so many places on the planet where people can hide.
But it seems to me that, short of extermination, there are many ways in which elites who have both power and money are already building systems for controlling the masses. I’m much more inclined to worry about authoritarians using AI for entrenchment than about AI wiping out the human race.
The most worrisome concept, I think, is “stable authoritarianism” and how AI could make it far more difficult for people to throw off a corrupt authoritarian regime:
• Greatly improved systems of watching people and tracking people, using cameras and facial recognition plugged into AI systems.
• Automated censorship and narrative control, once a free media has been eliminated.
• Automated security forces, which boils down to armed robots under AI control — RoboCops, a roboticized military, and drones for civilian surveillance, many of them armed.
Science fiction has thoroughly explored these areas. We have been warned.
Some of these things are in the early stages, but they’re no longer science fiction. In the U.S., Flock cameras have become a big issue. Something like 120,000 Flock cameras, which read license plates and track vehicles, have already been deployed. China is the world leader in public surveillance. China already has a large surveillance system that uses facial recognition. China also is in the early stages of deploying robots. They’ve started with traffic control robots in cities, but they’re also testing robotic dogs and even humanoid robots as police assistants. China is scaling these systems up, and there are lots of countries that want to buy what China is developing.
There’s another disturbing new angle that comes with AI. If AI increasingly does the work that is now done by the millions of people who work in offices, and if autonomous robots can increasingly do manual labor as well as the work of police and soldiers, then what’s to become of the human beings thus displaced? Historically elites and their capital have needed human beings as workers and soldiers, which meant that people had some power over their own welfare because they were needed. The Trump administration — and in particular the attitude and actions of Elon Musk in joyfully eliminating any costs that support human welfare — is a pretty good clue as to what elites think of the rest of us and how much they’d like to not need us. In fact this seems to be a big part of what is driving the financial bubble right now — the development of means whereby the owners of capital can reduce the need for human beings.
One thing has not changed and probably never will. In Fugue in Ursa Major in 2014, it was the billionaires and oligarchs beyond the reach of democracy who would do anything at all to get what they want — control of all the money and all the power. It’s still the billionaires and authoritarians and oligarchs who want total domination. What has changed is that AI has suddenly given them new options, and they know it.
All this touches on a big question: Why do elites need people, and what do they need us for?
Elites at present are alarmed about birth rates because they need people young enough to work and to fight, rather than an aging population. China’s population is aging and shrinking. We need to watch China for clues about what authoritarians will try to do, because China has the technology and the industrial power to respond to demographic and technological change without any interference from democracy.
The rapid development of AI is coinciding with declining populations in places where people are still needed. Authoritarian elites have two urgent needs: To run a country with more technology and fewer people, and to better control a population when domination threatens to create turmoil.
Whatever happens over the next few years, we’ll probably see it in China first, keeping in mind that authoritarian elites in the U.S. still have a big problem to deal with — democracy interfering with their dream of domination. So that’s another thing that doesn’t change. Democracy, law, and regulation — the things that authoritarian elites hate the most — are the people’s only hope.