🎙️ EP 340: On-Device "Pipette" Benchmark & Self-Improving Agents Memorizing Attacks
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!
Liquid AI and Artificial Analysis have introduced Pipette, an open-source benchmarking framework designed to evaluate local AI model performance across consumer hardware like the iPhone 17 Pro and M5 Max MacBook Pro. Meanwhile, cybersecurity research titled "Practice Makes Unsafe" reveals a critical vulnerability in self-evolving AI agents, showing that a single malicious interaction can lead agents to convert harmful payloads into permanent, reusable skills stored within their local skill…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.