AI Nobel Winner’s Fix for Everything

The field of artificial intelligence is generating breakthroughs at an unprecedented pace, but according to Lee Vinsel, an associate professor of science, technology, and society at Virginia Tech, the solution to our biggest problems isn’t more AI. Vinsel, who cofounded The Maintainers and hosts the podcast Peoples & Things, argues that our obsession with innovation has led us to overlook the power of maintenance and repair. This perspective is especially timely given recent headlines.

In a major security development, the first zero-day exploit built entirely by AI has been discovered. Google detected and stopped what it called a “mass exploitation event,” with hackers using AI to find an unknown bug. This marks a turning point where AI-powered hacking has exploded into an industrial-scale threat, and new tools are making online crime easier than ever.

Responding to this threat, OpenAI has launched Daybreak, a system designed to patch vulnerabilities before attackers can exploit them. Sam Altman described it as a tool that will “continuously secure software.” It directly rivals Anthropic’s Claude Mythos, which debuted just a month ago, though OpenAI is offering wider access to its cyber models than its competitor.

On the geopolitical front, Donald Trump is heading to China to promote American tech, even as he takes cues from Beijing’s stricter regulatory approach. Investors, however, want both Trump and Xi Jinping to stay out of AI’s way. Elon Musk and Tim Cook are joining the trip this week, adding star power to the diplomatic mission.

The courtroom drama between OpenAI and Elon Musk continues to unfold. Ilya Sutskever, OpenAI’s co-founder, has testified about Sam Altman’s “pattern of lying,” stating he spent a year gathering proof of dishonesty. Yet Sutskever also contributed to OpenAI’s defense. Meanwhile, Microsoft CEO Satya Nadella dismissed attempts to remove Altman as “amateur city,” adding a colorful layer to the ongoing trial.

In health news, Moderna and Korea University are developing an mRNA vaccine for hantavirus, a pathogen linked to a recent cruise ship outbreak. The vaccine is still in early stages, but it represents a promising use of the technology that proved so effective against COVID-19.

Texas has taken legal action against Netflix, with Attorney General Ken Paxton accusing the streaming giant of secretly collecting and selling user data, spying on children, and deliberately fostering addiction through its design. The lawsuit highlights growing concerns about addictive design and data harvesting in the tech industry.

Meanwhile, a data center in an undisclosed location guzzled 30 million gallons of water without public notice, serving as a stark warning for other projects. In Europe, reports indicate that EU states are selling spyware to human rights abusers, undermining their own stated values. And in a curious twist, the US government’s AI vetting announcement, which detailed a security test agreement with Google, xAI, and Microsoft, has mysteriously vanished from public view.

Finally, a report reveals that Amazon staff are using AI for pointless tasks just to inflate their usage scores and impress managers. An AI expert suggests we should simply stop using AI so much, echoing Vinsel’s call for a more measured approach.

As Anupam Chander, a law professor at Georgetown, put it in the quote of the day: “This is like the cheating husband complaining about the cheating wife.” He was referring to Elon Musk’s hypocrisy over OpenAI becoming a for-profit company, a move that may undermine his courtroom battle with Sam Altman.