The latest 404 Media podcast leads with a report from Lorenzo Franceschi-Bicchierai on how police can bypass the iPhone's inactivity reboot — a security feature designed to lock the device after a period of disuse. According to the episode, a leaked video claims to demonstrate this bypass, raising fresh concerns about law enforcement access to locked phones.
The same episode turns to a very different story: an online controversy over people 'torturing' large language models inside a simulated robot prison. The hosts describe the discourse as 'the dumbest debate in AI yet,' suggesting the outrage may be overblown but acknowledging the strangeness of the trend.
In the subscribers-only segment, the podcast discusses a New Mexico murder appeal where a lawyer cited fake witnesses invented by ChatGPT. The case underscores the risks of relying on AI-generated legal research without verification, a problem that has surfaced in multiple courtrooms recently.
While the three stories are unrelated, together they illustrate the widening societal impact of technology — from police tactics and AI safety debates to the legal system's struggle with generative AI errors.