Menu

Post image 1
Post image 2
Post image 3
1 / 3
0

This Week in AI: Rethinking the Agent Harness

Radar·Michelle Smith·3 days ago
#fptnr8qQ
#oreilly#agent#model#harness#john#experts
Reading 0:00
15s threshold

We kicked off our new weekly series This Week in AI on Monday, and we covered a lot of ground in 30 minutes, including an AI model that found security holes faster than decades of human auditing, a data center in Utah the size of two Manhattans, and a practical argument for why the harness you build around a model now matters more than which model you pick. Here are a few takeaways from the conversation between host Eric Freeman, faculty member at UT Austin and a longtime friend of O’Reilly , and guest John Berryman, founder of Arcturus Labs, an early production engineer on GitHub Copilot, and coauthor of O’Reilly’s Prompt Engineering for LLMs . Watch the entire episode to find out why you should be building your own agent and why John believes eventually there will be no internet for humans. AI’s security problem is now a policy problem You’ve probably already heard about Mythos .…

Continue reading — create a free account

Join HashtagPLUS to read full articles, follow hashtags, vote, and join the conversation.

Read More