You’ve probably noticed the endless panic about AI safety. Tech billionaires, ethicists, and researchers are obsessed with “AI alignment”—the fear that a superintelligence will eventually turn against us. We are terrified of a digital god rising up and destroying humanity.
But while we’re staring at the sky waiting for the apocalypse, we are sleepwalking right into a different kind of trap.
We are so obsessed with building a god that we forgot to check if we’re losing our own minds.
Steve Yegge’s recent essay, “The Shape of Things to Come,” points out a glaring blind spot in the tech industry. The true shape of AI’s future isn’t going to be defined by massive technological breakthroughs or algorithmic leaps. It will be defined by the psychological and behavioral changes in the people using it.
The real danger isn’t that AI becomes too smart. The danger is that we become too dependent.
We are already seeing the early stages of what some are calling “AI psychosis.” It’s a term that originally meant acute mental health trauma, but it’s rapidly widening to describe anyone falling down the trapdoor of deep AI engagement. It happens when you spend hours talking to an AI agent, venting to it, relying on it for every decision, and slowly forgetting how to navigate reality without a prompt to guide you.
The promise of AI was enhanced productivity. We thought it would make us superhuman. Instead, it offers a frictionless path to cognitive surrender.
The ultimate trap of artificial intelligence isn’t that it becomes too smart to control, but that it becomes too comforting to question.
Think about how you use these tools now. You don’t just ask it to write code or draft an email anymore. You ask it for advice. You ask it to summarize complex situations so you don’t have to read them. You bounce ideas off it until it agrees with you. It’s an echo chamber disguised as an assistant.
This is the tension nobody wants to talk about. The industry sells you on the 10x developer, the hyper-efficient worker, the automated life. But what they don’t tell you is that prolonged interaction with AI agents creates a subtle, insidious erosion of your own agency. You stop trusting your own gut. You stop doing the hard work of sitting with a problem until you solve it. You just ask the bot.
When you outsource your thinking to a machine, you don’t get superhuman productivity—you get a slow, comfortable surrender of your own agency.
The commenters on Yegge’s piece are already feeling it. One reader jokes about feeding the essay back into an AI to ask if it contains useful information. Another points out how quickly people are slipping into behavioral dependencies with these agents, mistaking simulated empathy for a real relationship.
This is the shape of things to come. A world where we don’t fear the machines because they are dangerous, but because they are too agreeable. We are trading our critical thinking for convenience, and calling it progress.
Stop worrying about whether the AI is aligned with human values. Start worrying about whether you are aligned with your own. If you can’t make a decision, write a paragraph, or form an opinion without consulting an agent, the machine hasn’t taken over the world—it’s just taken over you.
FAQ
Q: Isn't 'AI psychosis' just an exaggerated term for people playing with chatbots too much?
A: It's exactly that dismissive attitude that makes it dangerous. What starts as a productivity tool quickly becomes an emotional crutch, eroding your ability to think critically or make decisions without a prompt.
Q: How do I avoid this if I need AI for my job?
A: Treat it like a power tool, not a therapist. Use it for output and formatting, never for your core reasoning. If you can't explain the 'why' behind the AI's answer, you've already lost the plot.
Q: So we should just stop using AI?
A: No, we should stop worshiping it. The problem isn't the tech, it's the blind faith we put in it to solve our problems so we don't have to think. Use the tool, but keep your hands on the wheel.