KETTLE If you needed further evidence that AI comes first in pretty much everything nowadays, look no further than this year's Google Cloud Next show, which happened last week.

This week on the Kettle, systems editor Tobias Mann and cybersecurity editor Jessica Lyons join reporter Brandon Vigliarolo to talk not only about all the news out of GCN this past week, but also to discuss the other big AI headline: The incredibly small amount of time it took someone to gain improper access to Anthropic's Mythos cybersecurity AI model.

Google Cloud Next, ostensibly a cloud event, was packed with AI announcements like a split in Google's Tensor lineup with two versions of 8th generation chips coming out, one for inference and one for training. New AI security agents were also shown off, as was a refreshed version of Google's Vertex AI development platform, which has been renamed Gemini Enterprise Agent Platform and redesigned to serve as a single space to create and manage AI agents.

We also dive into the aforementioned Mythos leak and early Project Glasswing results, which suggest Mythos is, unsurprisingly, less capable than the hype made it out to be.

You can listen to The Kettle here, as well as on Spotify  Apple Music. ®

More about

• AI
• Anthropic
• Cybersecurity