Feb 6, 2023 · 35m · another-podcast

Generative search

Benedict Evans · 23m spoken Toni Cowan-Brown · 8m spoken
0:00 / 0:00

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

Toni Karen Brown and Benedict Evans discuss how generative AI is shifting software abstraction layers, disrupting traditional search into direct synthesis, and democratizing content creation while introducing critical ethical and legal challenges.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 73.1% of the talking time here. How this is scored →

The hosts as informed peer 7.6 Guest teaching 1.7 Guest disagreement 1.1 The hosts pushing back 1.9
05100:0010:0020:0030:000:00–2:07 · The hosts as informed peer 6/10 Framing the Generative AI Debate Beyond Cool Demos Evans immediately frames the episode around the 'so what' implications of generative AI rather than simple tech demos. The interaction is fully collaborative, with playful agreement on tech regulation topics.2:07–6:38 · The hosts as informed peer 8/10 How Interface Shifts and Abstractions Disrupt Established Software Evans draws deep historical parallels across computing paradigms including GUIs, cloud deployment, and mobile discontinuities. He explains how changing interface layers reset market inertia and create opportunities against incumbents like Google.6:38–11:49 · The hosts as informed peer 8/10 User Retention Dynamics and Enterprise Machine Learning Applications Brown shares her experience with conversational AI retention, and Evans responds with product analysis comparing Tinder, TikTok, and enterprise ML company Everlaw. He demonstrates mastery over how primitives get absorbed into vertical SaaS products.11:50–17:35 · The hosts as informed peer 8/10 The Creator Economy and Hyper-Personalized Generative Media When Brown posits that everyone is becoming an active creator, Evans pushes back with nuance by quoting DAU figures and historical TV ratings from 1960 to modern times. Both engage in a thoughtful discussion on content distribution and Mr. Beast.17:35–22:11 · The hosts as informed peer 7/10 Ethics of Deepfakes, Likeness Ownership, and Digital Identity Evans provides historical context on 19th-century photography and the evolution of image rights, while Brown educates him with a personal legal battle over an unauthorized Times Square billboard.22:11–30:49 · The hosts as informed peer 8/10 Redefining Search From Information Retrieval to Direct Generation Evans conceptualizes generative search beyond text into on-demand production pipelines like Shein and custom video generation. He weaves in references to early mobile networking and Douglas Adams' law of technology.30:49–35:26 · The hosts as informed peer 8/10 The Infinite Interns Paradigm and Emerging Moral Questions Evans outlines his mental model of generative AI as 'infinite interns' capable of generative execution rather than merely passive classification. Brown concludes with the looming unresolved questions of morality and digital ownership.0:00–2:07 · Guest teaching 1/10 Framing the Generative AI Debate Beyond Cool Demos Evans immediately frames the episode around the 'so what' implications of generative AI rather than simple tech demos. The interaction is fully collaborative, with playful agreement on tech regulation topics.2:07–6:38 · Guest teaching 1/10 How Interface Shifts and Abstractions Disrupt Established Software Evans draws deep historical parallels across computing paradigms including GUIs, cloud deployment, and mobile discontinuities. He explains how changing interface layers reset market inertia and create opportunities against incumbents like Google.6:38–11:49 · Guest teaching 2/10 User Retention Dynamics and Enterprise Machine Learning Applications Brown shares her experience with conversational AI retention, and Evans responds with product analysis comparing Tinder, TikTok, and enterprise ML company Everlaw. He demonstrates mastery over how primitives get absorbed into vertical SaaS products.11:50–17:35 · Guest teaching 2/10 The Creator Economy and Hyper-Personalized Generative Media When Brown posits that everyone is becoming an active creator, Evans pushes back with nuance by quoting DAU figures and historical TV ratings from 1960 to modern times. Both engage in a thoughtful discussion on content distribution and Mr. Beast.17:35–22:11 · Guest teaching 4/10 Ethics of Deepfakes, Likeness Ownership, and Digital Identity Evans provides historical context on 19th-century photography and the evolution of image rights, while Brown educates him with a personal legal battle over an unauthorized Times Square billboard.22:11–30:49 · Guest teaching 1/10 Redefining Search From Information Retrieval to Direct Generation Evans conceptualizes generative search beyond text into on-demand production pipelines like Shein and custom video generation. He weaves in references to early mobile networking and Douglas Adams' law of technology.30:49–35:26 · Guest teaching 1/10 The Infinite Interns Paradigm and Emerging Moral Questions Evans outlines his mental model of generative AI as 'infinite interns' capable of generative execution rather than merely passive classification. Brown concludes with the looming unresolved questions of morality and digital ownership.0:00–2:07 · Guest disagreement 1/10 Framing the Generative AI Debate Beyond Cool Demos Evans immediately frames the episode around the 'so what' implications of generative AI rather than simple tech demos. The interaction is fully collaborative, with playful agreement on tech regulation topics.2:07–6:38 · Guest disagreement 1/10 How Interface Shifts and Abstractions Disrupt Established Software Evans draws deep historical parallels across computing paradigms including GUIs, cloud deployment, and mobile discontinuities. He explains how changing interface layers reset market inertia and create opportunities against incumbents like Google.6:38–11:49 · Guest disagreement 1/10 User Retention Dynamics and Enterprise Machine Learning Applications Brown shares her experience with conversational AI retention, and Evans responds with product analysis comparing Tinder, TikTok, and enterprise ML company Everlaw. He demonstrates mastery over how primitives get absorbed into vertical SaaS products.11:50–17:35 · Guest disagreement 2/10 The Creator Economy and Hyper-Personalized Generative Media When Brown posits that everyone is becoming an active creator, Evans pushes back with nuance by quoting DAU figures and historical TV ratings from 1960 to modern times. Both engage in a thoughtful discussion on content distribution and Mr. Beast.17:35–22:11 · Guest disagreement 1/10 Ethics of Deepfakes, Likeness Ownership, and Digital Identity Evans provides historical context on 19th-century photography and the evolution of image rights, while Brown educates him with a personal legal battle over an unauthorized Times Square billboard.22:11–30:49 · Guest disagreement 1/10 Redefining Search From Information Retrieval to Direct Generation Evans conceptualizes generative search beyond text into on-demand production pipelines like Shein and custom video generation. He weaves in references to early mobile networking and Douglas Adams' law of technology.30:49–35:26 · Guest disagreement 1/10 The Infinite Interns Paradigm and Emerging Moral Questions Evans outlines his mental model of generative AI as 'infinite interns' capable of generative execution rather than merely passive classification. Brown concludes with the looming unresolved questions of morality and digital ownership.0:00–2:07 · The hosts pushing back 1/10 Framing the Generative AI Debate Beyond Cool Demos Evans immediately frames the episode around the 'so what' implications of generative AI rather than simple tech demos. The interaction is fully collaborative, with playful agreement on tech regulation topics.2:07–6:38 · The hosts pushing back 2/10 How Interface Shifts and Abstractions Disrupt Established Software Evans draws deep historical parallels across computing paradigms including GUIs, cloud deployment, and mobile discontinuities. He explains how changing interface layers reset market inertia and create opportunities against incumbents like Google.6:38–11:49 · The hosts pushing back 2/10 User Retention Dynamics and Enterprise Machine Learning Applications Brown shares her experience with conversational AI retention, and Evans responds with product analysis comparing Tinder, TikTok, and enterprise ML company Everlaw. He demonstrates mastery over how primitives get absorbed into vertical SaaS products.11:50–17:35 · The hosts pushing back 4/10 The Creator Economy and Hyper-Personalized Generative Media When Brown posits that everyone is becoming an active creator, Evans pushes back with nuance by quoting DAU figures and historical TV ratings from 1960 to modern times. Both engage in a thoughtful discussion on content distribution and Mr. Beast.17:35–22:11 · The hosts pushing back 1/10 Ethics of Deepfakes, Likeness Ownership, and Digital Identity Evans provides historical context on 19th-century photography and the evolution of image rights, while Brown educates him with a personal legal battle over an unauthorized Times Square billboard.22:11–30:49 · The hosts pushing back 2/10 Redefining Search From Information Retrieval to Direct Generation Evans conceptualizes generative search beyond text into on-demand production pipelines like Shein and custom video generation. He weaves in references to early mobile networking and Douglas Adams' law of technology.30:49–35:26 · The hosts pushing back 1/10 The Infinite Interns Paradigm and Emerging Moral Questions Evans outlines his mental model of generative AI as 'infinite interns' capable of generative execution rather than merely passive classification. Brown concludes with the looming unresolved questions of morality and digital ownership.

speaking balance: gold is the hosts, purple is the guest (3 minute bins)

0:00 · the hosts 57.9% · guest 42.1%0:00 · the hosts 57.9% · guest 42.1%3:00 · the hosts 94.6% · guest 5.4%3:00 · the hosts 94.6% · guest 5.4%6:00 · the hosts 64.7% · guest 35.3%6:00 · the hosts 64.7% · guest 35.3%9:00 · the hosts 78.5% · guest 21.5%9:00 · the hosts 78.5% · guest 21.5%12:00 · the hosts 55.3% · guest 44.7%12:00 · the hosts 55.3% · guest 44.7%15:00 · the hosts 69.7% · guest 30.3%15:00 · the hosts 69.7% · guest 30.3%18:00 · the hosts 72.7% · guest 27.3%18:00 · the hosts 72.7% · guest 27.3%21:00 · the hosts 68.8% · guest 31.2%21:00 · the hosts 68.8% · guest 31.2%24:00 · the hosts 86.9% · guest 13.1%24:00 · the hosts 86.9% · guest 13.1%27:00 · the hosts 81.9% · guest 18.1%27:00 · the hosts 81.9% · guest 18.1%30:00 · the hosts 63.5% · guest 36.5%30:00 · the hosts 63.5% · guest 36.5%33:00 · the hosts 84.6% · guest 15.4%33:00 · the hosts 84.6% · guest 15.4%
Sharpest disagreement ▶ 17:19 Brown insisting on active generation over passive consumption

Brown emphatically pushes her point that new generations are actively co-creating custom media crossovers rather than simply consuming preexisting broadcast media.

Hardest push from the hosts ▶ 12:44 Evans demurring on the definition of content creation

Evans openly questions Brown's claim that everyone is becoming a content creator, noting that posting family photos does not make someone a creator in the meaningful economic sense.

Biggest teaching moment ▶ 21:16 Brown's real-world unauthorized likeness battle

Brown brings concrete personal experience to the abstract discussion of image rights by detailing how she lost a legal battle when a brand put her image on a Times Square billboard without consent.

The host holds their own ▶ 14:50 Evans citing historical TV ratings share

Evans pulls up concrete broadcast audience share statistics from 1960 to 2020 to quantitatively illustrate media fragmentation and the unique reach of digital native creators like Mr. Beast.

the scores for every segment, with the reasoning behind each
ChapterTopicThe hosts as informed peerGuest teachingGuest disagreementThe hosts pushing backWhy
Framing the Generative AI Debate Beyond Cool Demos 6111 Evans immediately frames the episode around the 'so what' implications of generative AI rather than simple tech demos. The interaction is fully collaborative, with playful agreement on tech regulation topics.
How Interface Shifts and Abstractions Disrupt Established Software 8112 Evans draws deep historical parallels across computing paradigms including GUIs, cloud deployment, and mobile discontinuities. He explains how changing interface layers reset market inertia and create opportunities against incumbents like Google.
User Retention Dynamics and Enterprise Machine Learning Applications 8212 Brown shares her experience with conversational AI retention, and Evans responds with product analysis comparing Tinder, TikTok, and enterprise ML company Everlaw. He demonstrates mastery over how primitives get absorbed into vertical SaaS products.
The Creator Economy and Hyper-Personalized Generative Media 8224 When Brown posits that everyone is becoming an active creator, Evans pushes back with nuance by quoting DAU figures and historical TV ratings from 1960 to modern times. Both engage in a thoughtful discussion on content distribution and Mr. Beast.
Ethics of Deepfakes, Likeness Ownership, and Digital Identity 7411 Evans provides historical context on 19th-century photography and the evolution of image rights, while Brown educates him with a personal legal battle over an unauthorized Times Square billboard.
Redefining Search From Information Retrieval to Direct Generation 8112 Evans conceptualizes generative search beyond text into on-demand production pipelines like Shein and custom video generation. He weaves in references to early mobile networking and Douglas Adams' law of technology.
The Infinite Interns Paradigm and Emerging Moral Questions 8111 Evans outlines his mental model of generative AI as 'infinite interns' capable of generative execution rather than merely passive classification. Brown concludes with the looming unresolved questions of morality and digital ownership.

Statements from this episode (10)

Insight
Evans: UX and platform shifts create discontinuities that produce breakout companies
“Just shifting the UX and the platform is often kind of a discontinuity that creates new companies. And so you have that when you go to cloud, you get Salesforce. When you go to mobile, you get not just Instagram, but also Snapchat, and then TikTok”
Benedict Evans Feb 6, 2023 ▶ 5:05
Opinion
Evans: A novel search interface could seize market share from incumbents
“Simply doing a new search engine with a completely new user interface could be an opportunity to take share, even if it's actually just a Up front and onto the same thing you had before.”
Benedict Evans Feb 6, 2023 ▶ 6:26
Insight
Evans: TikTok differs from YouTube in UX, not fundamental product capability
“TikTok is not Different to YouTube in any sort of fundamental product sense. It's different in UX.”
Benedict Evans Feb 6, 2023 ▶ 8:57
Opinion
Evans: MrBeast transforms TV more fundamentally than Netflix
“Mr. Beast is changing what TV is in a way that Netflix is not. You know, Netflix is a TV company. It's a TV company with a different distribution model. Mr. Beast is, is, is a much more fundamental change in what TV means than Netflix paying TV people to make …”
Benedict Evans Feb 6, 2023 ▶ 14:03
Assertion Supported
Evans: Top US TV show reach dropped from 40% in 1960 to 10% in 2020
“So, 1960 Ratings for the top-ranked US show was about 40% of US TV households, and at that point, everyone in America has a TV, like 90, 95% of households have a TV, so basically 40% of Americans, whatever the top show was that year, 40% of Americans saw it, a…”
Benedict Evans Feb 6, 2023 ▶ 15:28
Prediction Not checkable as stated
Evans: Kids' TV voiceovers and animation will be automated by computers
“It's Very easy to look at this and think, okay, all those voiceovers are going to be done by computer next. And then all the animation is going to be done by computer as well.”
Benedict Evans Feb 6, 2023 ▶ 16:41
Insight
Evans: Dismissing generative AI over early errors misses the conceptual breakthrough
“And clearly you should be more, if you look at it and you say, well, it makes cats, but the cats have four, have three legs. If you say, well, yeah, it can make some text, but the text isn't very good and it's got mistakes in it. You're frankly, you're an idio…”
Benedict Evans Feb 6, 2023 ▶ 26:21
Prediction Held up
Evans: Custom video effects and actor swaps will run on iPhone apps
“Now, what, now, I, well, what is the point in which I say, okay, I would actually like to see that, but I would like to see Don Rickles instead of Go Forever. That, that is not 20 people in a production studio and a whole pile of SFX. That is going to be an iP…”
Benedict Evans Feb 6, 2023 ▶ 27:45
Insight
Evans: Machine learning acts as 'infinite interns' for simple tasks
“Well, the analogy I always used to use talking about machine learning was that this gives you infinite intents, that all these sort of things where it doesn't need an expert, you just, you know, because I'm listening to the call, coming into the call center, i…”
Benedict Evans Feb 6, 2023 ▶ 32:59
Assertion Not checkable as stated
Evans: Recasting Star Wars with Burt Reynolds via AI is a weekend project
“Reshoot, give me the complete original Star Wars movie with Burt Reynolds instead of Harrison Ford, and that's a weekend project.”
Benedict Evans Feb 6, 2023 ▶ 34:47
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 100 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.