Wired podcast weighs Anthropic AI warning, Apple foldable iPhone, and census data

In short: A WIRED podcast episode looks at three separate stories that all raise questions about trust, from AI safety claims to government data and new Apple devices.

What happened

WIRED’s podcast Uncanny Valley discussed a viral resignation by Jacob Coxon, a former AI researcher at Anthropic. Anthropic is an AI company known for its chatbot Claude. Coxon said there is a real chance that AI could wipe out humanity within a decade, and the claim spread widely online.

In the episode, WIRED reporter Will Knight said these fears are not new inside leading AI labs. He also argued that people often confuse what is possible with what is likely. He said a more immediate worry is companies rushing AI tools into the real world, where they can fail in surprising ways, like an unreliable helper that still gets put in charge.

The same episode also covered Apple’s latest product event. WIRED said Apple announced a foldable iPhone called the iPhone Duo, priced around $2,000. The show also discussed new Apple Watch features that can listen for sounds and create short transcripts of nearby speech, which raised privacy concerns because people around the watch may not know it is enabled.

Finally, the podcast talked about a WIRED investigation into a US Census Bureau report that claimed noncitizens voted in the 2020 election. WIRED reported that the analysis relied on faulty data matching, and that the report’s authors were not listed, which experts said is unusual.

Why it matters

All three topics come back to the same basic issue, which is whether the public can trust powerful systems. That includes AI built by a few companies, devices that can capture what people say, and official government reports that may shape future policy.

Source: Wired