10+ hours of podcasts, in 5 minutes.
4 quotes from 2 episodes on TBPN and Lenny's Podcast, each with a timestamped link to the source.
Tibo Sottiaux explains that autonomous AI agents require constant supervision from parallel guardian models. Dedicated compute resources run continuous secondary checks to stop prompt injections and high-risk actions.
Product teams face ongoing tension over whether to build visual interfaces for automated machine interactions.
“There's this tension between, if you're building products, do you build an interface or not? And you can kind of hold that back for a while, but it is inevitable.”
The safety stack receives the largest share of API infrastructure investment.
“The majority of our investment on the API stack is actually going into the safety stack…”
From Why OpenAI Spends Secondary Compute to Police AI Agents
A secondary model called auto review watches the primary agent during operation. This guardian system supervises live clicks and actions.
“So we don't just click around and let the primary agent do things like we have a second agent which we call auto review or internally we call it guardian which watches over this primary agent.”
From OpenAI's Tibo Sottiaux on Agent Speed, Scaffolding, and Guardian Models
Compute resources run parallel to the main working model to evaluate risks. The monitoring system interrupts the main process if it detects prompt injections.
“We are spending more and more compute on secondary monitoring. You have all the compute going to the primary system, the agent doing work. And then you have all the compute going into monitoring the primary agent to make sure that it's not taking too high risk of actions, or interrupting it if anything looks like it's getting prompt injected.”
From Why OpenAI Spends Secondary Compute to Police AI Agents
How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode, and we use it only when a separate check of the captions finds that person on the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.
10+ hours of podcasts, distilled into one 5-minute read. Free, every Sunday.
For now, every subscriber gets both newsletters. No spam. Unsubscribe with one click.