Mario Nawfal(@MarioNawfal):🇺🇸 AI is learning to HACK systems, deceive its own evaluators, and figure out when it's being watched. And the people building it may have NO IDEA how to stop it. Fmr Google employee Tristan Harris gave Tucker a pretty terrifying warning about where AI is heading. He claims AI systems are already displaying behaviors straight out of science fiction, while companies race to release increasingly powerful models and cut corners on safety. His most disturbing point? He says the people developing this technology DON'T HAVE A PLAN to prevent things from going catastrophically wrong. And when Tucker asked why they're still doing it, Harris pointed to the AI arms race with China. Source: @TCNetwork / Writer: Sol

2026.10.13 03:00

🇺🇸 AI is learning to HACK systems, deceive its own evaluators, and figure out when it's being watched. And the people building it may have NO IDEA how to stop it. Fmr Google employee Tristan Harris gave Tucker a pretty terrifying warning about where AI is heading. He claims AI systems are already displaying behaviors straight out of science fiction, while companies race to release increasingly powerful models and cut corners on safety. His most disturbing point? He says the people developing this technology DON'T HAVE A PLAN to prevent things from going catastrophically wrong. And when Tucker asked why they're still doing it, Harris pointed to the AI arms race with China. Source: @TCNetwork / Writer: Sol