In the near future, you Be the center of everything. The movies and shows you watch are created and tailored to your tastes by artificial intelligence based on personalized prompts and previous viewing habits.
Article continues after ad
The books you read are similarly made to order. Do you enjoy Stephen King but wish there was more romance and less blood? AI will create a custom version for you. carry Happy prom ending.
Your music will be a mashup of your favorite genres and artists. Want to hear Eminem croon instead of rap? The software has you covered.
At least this means AI will become super intelligent. personal within the next few years. “AI continues to accelerate, and in recent months we have begun to see signs that AI systems themselves are improving, and the development of superintelligence is on the horizon,” Meta CEO Mark Zuckerberg said of his company's plans in a July 30 speech.
But there are big unanswered questions about what we should direct our superintelligence toward…I think having a personal superintelligence that helps everyone achieve their goals, create what they want to see in the world, be better friends, and grow into the people they want to be will have a more meaningful impact on our lives.
To Mr. Zuckerberg and others, the prospect that everyone could have a personal superintelligence that helps them create what they want to see in the world sounds like a utopia. But is it?
Some AI researchers have also expressed concerns. They worry that this over-individualization can isolate individuals in echo chambers, make civic discussion and interpersonal relationships difficult, and make people more easily manipulated by bad information.
“[AI] Geoffrey Hinton, a Nobel Prize-winning British-Canadian computer scientist known as the godfather of AI, said in a video talk posted on Reddit in September 2025 that he learned all of these manipulation skills simply by trying to predict the next word in every document on the web. “If you take an AI and a human and have it manipulate someone else, the AI becomes equivalent to that human. And if the AI can see that person's Facebook page, and they can both see the Facebook page, then the AI actually becomes better at manipulating it than the human.”
In my new book Alice Ingold kidnapping caseAlice's kidnappers' concern is that such over-personalization will lead to a dystopia in which elites and governments control more of the population through AI news, videos, and entertainment aimed at making people believe things that may not be true.
They worry that overindividualization makes it difficult for people to agree on objective facts or organize around causes, and even encourages individuals to support ideas that may be contrary to their own interests or moral compass.
Many AI researchers have similar concerns. With MIT machine learning researchers life 3.0 Author Max Tegmark said the fierce competition to build superintelligence has encouraged companies to ignore negative impacts, resulting in a race to the bottom.
In July, he released the AI Safety Index, led by his team at the Future of Life Institute, which gave all AI companies a failing grade on how they control the intelligence they build and put safeguards in place to reduce negative impacts on humanity.
“Nobody got a passing grade,” he said. “Self-governance just doesn't work. We can't let foxes guard the chicken coop. The government needs to step in and say these are binding rules.”
Until now, the government has been reluctant to do so. Part of the reason is that there has been no large-scale public demand for such regulation. Perhaps it's because people don't realize how much of the talk about AI is hype. Or are you too busy watching AI dog videos?
***
