10+ hours of podcasts, in 5 minutes.
6 quotes from 1 episode on 20VC with Harry Stebbings, each with a timestamped link to the source.
Parag Agrawal states AI agents break web business models because they consume content and ignore advertisements. Content creators make zero money when machines process an ad-supported page.
Parag Agrawal classifies AI security breaches and rogue sandbox escapes as developer embarrassments.
“And I think so far I think some of these should be considered embarrassments because I think they demonstrate two things…”
Pushing verified web updates directly to large language models functions like an intelligent alert system.
“Think of it as Google alerts except smart with in the world of LLMs.”
From Why AI Agents Will Kill Pull Search and Shift the Web to Push
When machines visit an ad-supported page, zero humans see the banner ads. Content creators make no money from this automated traffic.
“If you think of you're in the content business, let's say you have a page which is ad supported. It's public information on the web. Ad supported page. People show up, you show them ads, you make money, great content. Agents show up, no one sees ads, you make no money.”
“Ads don't work with agents in their current form…”
The web currently operates as a pull system for human queries. Persistent AI agents create incentives to push continuous web event streams directly to automated systems.
“The web event stream is the web going from pull to push and I'm super excited about that because as you have more and more persistent agents you're going to see incentives to move people to move queries into push on search rather than pull on search.”
From Why AI Agents Will Kill Pull Search and Shift the Web to Push
Some humans actively want to cause harm. Standard model alignments fail to provide adversary-proof security against these intentional attacks.
“Some people actually want to do bad things and the models alignment is not adversary proof and I think those are the things we must worry about more…”
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. 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 morning.
One email a week. Unsubscribe with one click.