StarZero All shows Try the podcast clipper →

Feb 22, 2024 · 28m · how-i-built-this

The peril (and promise) of AI with Tristan Harris: Part 1

Tristan Harris · 16m spoken Guy Raz · 9m spoken
0:00 / 0:00

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

In this episode of How I Built This Lab, Guy Raz interviews Center for Humane Technology co-founder Tristan Harris about how perverse engagement incentives led from the attention economy crisis to today's unconstrained, exponential race in generative AI.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Guy holds 36% of the talking time here. How this is scored →

Guy as informed peer 3.8 Guest teaching 3.8 Guest disagreement 1.0 Guy pushing back 0.0
05100:0010:0020:002:31–4:50 · Guy as informed peer 3/10 Lessons from Apture: How Incentives Shape Tech Raz opens by asking Harris to describe the early product mechanics of Apture. Harris elaborates on how investor demands shifted their metric from educational enrichment to ad impressions, establishing the core theme of incentive structures.4:51–10:11 · Guy as informed peer 4/10 The Google Manifesto on the Attention Economy Raz quotes directly from Harris's internal Google manifesto and connects the attention crisis to Neil Postman and Aldous Huxley. Harris explains how smartphones expanded the finite boundaries of human attention into a commercial commodity.10:12–17:40 · Guy as informed peer 4/10 Mid-Episode Break: Teasing the Pace of Tech Growth Raz provides an extended historical comparison to the Gutenberg press and the Protestant Reformation to illustrate societal adaptation timelines. Harris builds on this by citing E.O. Wilson and introducing the concept of double exponential growth in AI development.17:41–27:44 · Guy as informed peer 4/10 Mid-Episode Break: The AI Threat Teaser Raz asks about the 2017 transformer breakthrough and the Manhattan Project comparison. Harris explains how scaling compute and data creates unpredicted emergent capabilities like humor analysis and chemical synthesis.2:31–4:50 · Guest teaching 3/10 Lessons from Apture: How Incentives Shape Tech Raz opens by asking Harris to describe the early product mechanics of Apture. Harris elaborates on how investor demands shifted their metric from educational enrichment to ad impressions, establishing the core theme of incentive structures.4:51–10:11 · Guest teaching 3/10 The Google Manifesto on the Attention Economy Raz quotes directly from Harris's internal Google manifesto and connects the attention crisis to Neil Postman and Aldous Huxley. Harris explains how smartphones expanded the finite boundaries of human attention into a commercial commodity.10:12–17:40 · Guest teaching 4/10 Mid-Episode Break: Teasing the Pace of Tech Growth Raz provides an extended historical comparison to the Gutenberg press and the Protestant Reformation to illustrate societal adaptation timelines. Harris builds on this by citing E.O. Wilson and introducing the concept of double exponential growth in AI development.17:41–27:44 · Guest teaching 5/10 Mid-Episode Break: The AI Threat Teaser Raz asks about the 2017 transformer breakthrough and the Manhattan Project comparison. Harris explains how scaling compute and data creates unpredicted emergent capabilities like humor analysis and chemical synthesis.2:31–4:50 · Guest disagreement 1/10 Lessons from Apture: How Incentives Shape Tech Raz opens by asking Harris to describe the early product mechanics of Apture. Harris elaborates on how investor demands shifted their metric from educational enrichment to ad impressions, establishing the core theme of incentive structures.4:51–10:11 · Guest disagreement 1/10 The Google Manifesto on the Attention Economy Raz quotes directly from Harris's internal Google manifesto and connects the attention crisis to Neil Postman and Aldous Huxley. Harris explains how smartphones expanded the finite boundaries of human attention into a commercial commodity.10:12–17:40 · Guest disagreement 1/10 Mid-Episode Break: Teasing the Pace of Tech Growth Raz provides an extended historical comparison to the Gutenberg press and the Protestant Reformation to illustrate societal adaptation timelines. Harris builds on this by citing E.O. Wilson and introducing the concept of double exponential growth in AI development.17:41–27:44 · Guest disagreement 1/10 Mid-Episode Break: The AI Threat Teaser Raz asks about the 2017 transformer breakthrough and the Manhattan Project comparison. Harris explains how scaling compute and data creates unpredicted emergent capabilities like humor analysis and chemical synthesis.2:31–4:50 · Guy pushing back 0/10 Lessons from Apture: How Incentives Shape Tech Raz opens by asking Harris to describe the early product mechanics of Apture. Harris elaborates on how investor demands shifted their metric from educational enrichment to ad impressions, establishing the core theme of incentive structures.4:51–10:11 · Guy pushing back 0/10 The Google Manifesto on the Attention Economy Raz quotes directly from Harris's internal Google manifesto and connects the attention crisis to Neil Postman and Aldous Huxley. Harris explains how smartphones expanded the finite boundaries of human attention into a commercial commodity.10:12–17:40 · Guy pushing back 0/10 Mid-Episode Break: Teasing the Pace of Tech Growth Raz provides an extended historical comparison to the Gutenberg press and the Protestant Reformation to illustrate societal adaptation timelines. Harris builds on this by citing E.O. Wilson and introducing the concept of double exponential growth in AI development.17:41–27:44 · Guy pushing back 0/10 Mid-Episode Break: The AI Threat Teaser Raz asks about the 2017 transformer breakthrough and the Manhattan Project comparison. Harris explains how scaling compute and data creates unpredicted emergent capabilities like humor analysis and chemical synthesis.

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

0:00 · Guy 74.7% · guest 25.3%0:00 · Guy 74.7% · guest 25.3%3:00 · Guy 22.9% · guest 77.1%3:00 · Guy 22.9% · guest 77.1%6:00 · Guy 25.6% · guest 74.4%6:00 · Guy 25.6% · guest 74.4%9:00 · Guy 54.2% · guest 45.8%9:00 · Guy 54.2% · guest 45.8%12:00 · Guy 44% · guest 56%12:00 · Guy 44% · guest 56%15:00 · Guy 13.5% · guest 86.5%15:00 · Guy 13.5% · guest 86.5%18:00 · Guy 37.8% · guest 62.2%18:00 · Guy 37.8% · guest 62.2%21:00 · Guy 22.5% · guest 77.5%21:00 · Guy 22.5% · guest 77.5%24:00 · Guy 19.9% · guest 80.1%24:00 · Guy 19.9% · guest 80.1%27:00 · Guy 56.2% · guest 43.8%27:00 · Guy 56.2% · guest 43.8%
Sharpest disagreement ▶ 24:36 Dismissing early AI alarmism

Harris forcefully pushes back against the conventional Effective Altruism AI doom framing, recounting how he told them they were completely misguided for missing the immediate threat of social media.

 Hardest push from Guy ▶ 7:57 Probing Google leadership sincerity

Raz challenges whether Google leadership genuinely engaged with Harris's manifesto or merely offered cynical lip service before sidelining the issue.

 Biggest teaching moment ▶ 26:10 Explaining unexpected emergent abilities

Harris explains the technical nature of transformer models, breaking down how scaling compute creates unpredictable emergent capabilities that even the model creators did not anticipate.

 Guy holds their own ▶ 14:04 Historical analysis of the printing press

Raz demonstrates deep contextual knowledge by synthesizing 500 years of post-Gutenberg European history to frame the psychological shock of rapid information proliferation.
the scores for every segment, with the reasoning behind each
ChapterTopicGuy as informed peerGuest teachingGuest disagreementGuy pushing backWhy
Lessons from Apture: How Incentives Shape Tech 3310 Raz opens by asking Harris to describe the early product mechanics of Apture. Harris elaborates on how investor demands shifted their metric from educational enrichment to ad impressions, establishing the core theme of incentive structures.
The Google Manifesto on the Attention Economy 4310 Raz quotes directly from Harris's internal Google manifesto and connects the attention crisis to Neil Postman and Aldous Huxley. Harris explains how smartphones expanded the finite boundaries of human attention into a commercial commodity.
Mid-Episode Break: Teasing the Pace of Tech Growth 4410 Raz provides an extended historical comparison to the Gutenberg press and the Protestant Reformation to illustrate societal adaptation timelines. Harris builds on this by citing E.O. Wilson and introducing the concept of double exponential growth in AI development.
Mid-Episode Break: The AI Threat Teaser 4510 Raz asks about the 2017 transformer breakthrough and the Manhattan Project comparison. Harris explains how scaling compute and data creates unpredicted emergent capabilities like humor analysis and chemical synthesis.

Statements from this episode (10)

Opinion
Technology Cannot Serve Humanity Without Fixing Underlying Structural Incentives
“We only do that because we care about the vision where technology actually is in service of humanity. And to do that, we just think we have to fix the incentives.”
Tristan Harris Feb 22, 2024 ▶ 2:24
Assertion Not checkable as stated
Major Publishers Evaluated Apture Solely on Dwell Time and Ad Impressions
“We did deals with the Washington post, the New York times, the economist, and they would ask us one question, which was how much did you increase the amount of time people spent on my website? Okay. How many ad views impressions did you increase?”
Tristan Harris Feb 22, 2024 ▶ 3:52
Insight
Tech Companies Cannot Sustain Infinite Growth on Finite Human Attention
“You can't run infinite growth of every company demanding more attention per quarter per year forever. With a finite supply of human attention.”
Tristan Harris Feb 22, 2024 ▶ 6:02
Prediction Not checkable as stated
Combining AI With Attention Incentives Will Exacerbate Societal Information Crises
“It's the big incentives of the attention economy that have made this problem worse, and then combine that with technology and AI. And we're heading into, you know, a bigger version of that.”
Tristan Harris Feb 22, 2024 ▶ 9:48
Insight
Google, Android, and Chrome Inherently Benefit From Maximizing User Screen Time
“At the end of the day, Android and Chrome, you know, and Google benefit the more overall screen time there is. Versus not.”
Tristan Harris Feb 22, 2024 ▶ 12:18
Insight
AI Creates a Double Exponential Curve by Accelerating Its Own Development
“Then you get AI and you get a double exponential. You get an exponential that accelerates its own exponential, meaning AI accelerates AI development.”
Tristan Harris Feb 22, 2024 ▶ 15:56
Assertion Supported
ChatGPT Reached 100 Million Users in Two Months Compared to Instagram
“It took Instagram, I think two years to get to a hundred million users and it took chat GPT two months.”
Tristan Harris Feb 22, 2024 ▶ 16:23
Assertion Partly supported
OpenAI, Anthropic, and DeepMind Were Founded Expressly to Create AGI
“Remember that the stated mission statement of OpenAI, Anthropic, and DeepMind from the very beginning, those three companies, Have been to create artificial general intelligence.”
Tristan Harris Feb 22, 2024 ▶ 20:08
Insight
Scaling Neural Networks Produces Emergent Capabilities Unanticipated by Creators
“And the thing about it is that as you scale these digital brains, they end up with more emergent capabilities that no one who built them anticipated.”
Tristan Harris Feb 22, 2024 ▶ 26:02
Assertion Contradicted
Researchers Took Two Years to Discover GPT-3.5 Could Synthesize Dangerous Chemicals
“It took two years, for example, for researchers to figure out that GPT-III.V, the predecessor to GPT-IV, had research-grade chemistry knowledge, meaning you could ask it questions about, you know, how to synthesize dangerous chemicals, and it would tell you ho…”
Tristan Harris Feb 22, 2024 ▶ 26:50
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 800 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.

Try the podcast clipper Learn more