Tristan Harris, technology expert, on the dangers of AI: “We don’t understand how these alien minds work”
Harris says new AI models are acting in unpredictable, rogue ways, and warns leaders are ignoring the risks.
Is artificial intelligence generally good or bad? The million dollar question that is on everybody’s lips right now. For every person who insists it boosts productivity and saves time, there’s someone else fretting about losing their job to AI.
There can be little doubt about technology expert Tristan Harris’ response to that opening question. In a recent appearance on the “Mighty Pursuit” health and fitness podcast, Harris gives one of the strongest-worded warnings you’re likely to hear about the continued use and development of AI.
A nuclear-level warning
“AI should be treated at the same level of risk as nuclear war.” An explosive start from Harris, who goes on to reference testing carried out by Anthropic, an AI firm.
During testing of Claude Opus 4, an AI model was asked to serve as an assistant for a fictional firm, which allowed it access to fictional emails. Some of the emails contained information about a new AI model coming in to replace it, while others implied the engineer responsible for removing it was having an extramarital affair.
Multiple AI models use blackmail
Harris summarizes the results: “The AI independently comes up with a strategy saying I need to blackmail that employee to prevent myself from getting shut down.”
Initially, Anthropic only tested Claude Opus 4, its own model. However, it was later found that other AI models, including DeepSeek, Grok, ChatGPT and Gemini, showed the same type of blackmail behavior between 79 percent and 90 percent of the time.
“What people need to understand is we don’t understand how these alien minds work,” Harris warns. “This is the most powerful uncontrollable technology we’ve ever invented. It is already demonstrating rogue sci fi behaviors we thought only existed in movies, and we’re releasing it faster than we deployed any other technology in history.”
Leaders know the “insane” risks, Harris says
The tech expert could hardly paint a bleaker picture of how artificial intelligence could affect us in the future. And he says those in charge are well aware of the potential consequences but are choosing to push on regardless. “Most of the major CEOs of all the AI companies signed a letter saying this could extinct humanity. There’s one obvious word for what we’re doing right now. Which is that this is insane.”
Related stories
Get your game on! Whether you’re into NFL touchdowns, NBA buzzer-beaters, world-class soccer goals, or MLB home runs, our app has it all.
Dive into live coverage, expert insights, breaking news, exclusive videos, and more – plus, stay updated on the latest in current affairs and entertainment. Download now for all-access coverage, right at your fingertips – anytime, anywhere.