Ilya Sutskever Testifies in Musk v. Altman Trial: "I Didn't Want OpenAI to Be Destroyed"

Ilya Sutskever, OpenAI's former chief scientist, took the stand on May 11 in the Musk v. Altman trial. Addressing his own role in the November 2023 ouster of Sam Altman, he justified his actions by saying he "did not want OpenAI to be destroyed."

Sutskever has since left OpenAI to found his own venture, SSI (Safe Superintelligence Inc.), but he appeared in court as a defense witness for his former employer. That he — a former chief scientist who has distanced himself from the company — still took the stand on OpenAI's behalf underscores the tangled personal dynamics at play in this trial.

The lawsuit was brought by Elon Musk over OpenAI's departure from its nonprofit roots, and the proceedings are examining a fundamental conflict over governance and direction in the AI industry.

Linux Hit by Second Critical Vulnerability in Two Weeks: "Dirty Frag" Enables Root Takeover from Containers

Yet another critical vulnerability has been discovered in Linux. Hot on the heels of last week's disclosure of a threat dubbed "Copy Fail," a newly identified vulnerability named "Dirty Frag" has been found to allow root privileges to be seized from low-privilege users and container environments.

Microsoft has already reported signs that attackers are experimenting with the vulnerability in real-world environments. The leaked exploit code works reliably and doesn't even cause crashes, making it highly stealthy. The risk is especially acute in shared server environments, and the vulnerability could have a broad impact on AI inference infrastructure and cloud platforms. Production patches are being rolled out in stages, and prompt application is recommended.

GM Lays Off Hundreds of IT Employees in Move to Replace Them with AI Talent

General Motors (GM) has reportedly laid off hundreds of IT employees as part of a plan to replace them with workers who have AI skills. The move, which drew attention through TechCrunch's reporting and discussion on Hacker News, has sparked debate as a concrete example of corporate AI shifts starting to hit employees directly.

As demand for "AI skills" surges across the tech industry, some warn that existing IT departments could be phased out. At the same time, the supply of AI talent continues to fall short of demand, so whether a simple personnel swap will actually work remains an open question.

Mistral AI's NPM Package Compromised, Highlighting Supply Chain Security Challenges

The NPM package for French AI startup Mistral AI's TypeScript client has been found to be compromised. The incident, reported via a GitHub issue, has once again raised concerns about the supply chain security of AI tools.

With the use of AI-related packages exploding, ensuring the safety of dependencies has become a challenge for the entire developer community.

Robinhood Rides the AI Investment Boom, Prepares Second Venture Fund

Riding the AI investment boom, Robinhood is preparing to launch its second venture fund. The firm plans to invest not only in early-stage startups but also in growth-stage companies, a sign that appetite for AI investments is spreading to retail-focused platforms as well.

Wrap-Up

Today's AI news brought significant developments along three axes: technology, security, and employment. Sutskever's testimony strikes at the heart of the AI governance debate, the back-to-back Linux vulnerabilities sound an alarm over the security of AI infrastructure, and GM's layoffs exemplify the social impact of the AI shift. Alongside technological progress, the security of the infrastructure underpinning it and the movement of talent look set to be key themes going forward.