Skip to content
All-In PodcastAll-In Podcast

E168: Can Google save itself? Abolish HR, AI takes over Customer Support, Reddit IPO teardown

(0:00) Bestie intros! Jato Caelin (1:36) Groq update!: The key metric developer platforms should care about (5:55) Google's make or break moment: Should Sundar be fired, AI safety team's power, RIF potential, stock pressure, and more (29:17) How bloated HR departments slow down companies (40:15) Google's licensing deals: Is this TAC 2.0? How can Reddit capitalize on its data? (53:50) Klarna's huge AI innovation: AI disruption, open-source benefits, a new class of billion-dollar startups (1:13:06) Reddit S-1 breakdown (1:22:52) The Apple Car is dead: Apple cancels Project Titan Follow the besties: https://twitter.com/chamath https://twitter.com/Jason https://twitter.com/DavidSacks https://twitter.com/friedberg Follow the pod: https://twitter.com/theallinpod https://linktr.ee/allinpodcast Intro Music Credit: https://rb.gy/tppkzl https://twitter.com/yung_spielburg Intro Video Credit: https://twitter.com/TheZachEffect Referenced in the show: https://www.google.com/finance/quote/GOOG:NASDAQ https://twitter.com/PirateWires/status/1762842476744708171 https://twitter.com/lulumeservey/status/1762864255437516932 https://x.com/pmarca/status/1762532979317043515 https://twitter.com/shellenberger/status/1762900581134557636 https://twitter.com/shaunmmaguire/status/1760885099984261564 https://www.reuters.com/technology/reddit-ai-content-licensing-deal-with-google-sources-say-2024-02-22 https://techcrunch.com/2024/02/29/google-brings-stack-overflows-knowledge-base-to-gemini https://www.sec.gov/Archives/edgar/data/1713445/000162828024006294/reddits-1q423.htm https://static01.nyt.com/newsgraphics/documenttools/82a013b9ba852548/9d4b1790-full.pdf https://www.klarna.com/international/press/klarna-ai-assistant-handles-two-thirds-of-customer-service-chats-in-its-first-month https://www.google.com/finance/quote/TLPFY:OTCMKTS https://tech.facebook.com/artificial-intelligence/2022/10/ai-translation-unwritten-language https://www.engadget.com/metas-new-multimodal-translator-uses-a-single-model-to-speak-100-languages-133040214.html https://www.youtube.com/shorts/hv1HzqwcvTA https://www.youtube.com/watch?v=e-NNuZJrrcY https://www.youtube.com/shorts/c6mimFksLBg https://twitter.com/balajis/status/1763067849218850939 https://old.reddit.com/r/AskReddit/comments/3cs78i/whats_the_best_long_con_you_ever_pulled/cszjqg2 https://www.bloomberg.com/news/articles/2024-02-27/apple-cancels-work-on-electric-car-shifts-team-to-generative-ai #allin #tech #news

Chamath PalihapitiyahostJason CalacanishostDavid Friedberghost
Feb 29, 20241h 26mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Google’s AI crisis, DEI backlash, and AI-powered customer support shift

  1. The hosts dissect Google’s Gemini image-generation debacle as a symptom of deeper cultural and structural issues at Google, especially the power of DEI/“Responsible AI” and HR over product decisions, and debate whether founders Larry and Sergey or new leadership can still save the company in AI.
  2. They then explore the emerging market for AI training data, including Google’s multi‑million‑dollar deals with Reddit and Stack Overflow, and whether this resembles a new “TAC 2.0” or more traditional content licensing, with unclear long‑term economics.
  3. Klarna’s AI assistant replacing the work of 700 customer support agents is highlighted as an early concrete example of AI driving major cost savings, with discussion of job displacement, new types of work, and the threat to call-center and SaaS businesses.
  4. Finally, they break down Reddit’s IPO S‑1, questioning its growth and low ARPU, and touch on Apple’s reported cancellation of its decade‑long car project in favor of refocusing on generative AI.

IDEAS WORTH REMEMBERING

5 ideas

Google’s Gemini incident reflects a structural DEI/HR problem, not a one-off bug.

The panel argues that Gemini’s racially skewed image outputs were the predictable result of empowered DEI/Responsible AI functions and a monocultural leadership environment, where dissent was socially and professionally punished, leading to distorted products.

Founders and boards must reassert product-focused, meritocratic leadership to stay competitive in AI.

Comparing Google to Meta and X/Twitter, they contend that only strong founder-style intervention—replacing leaders, shrinking overgrown bureaucracies, and empowering product owners over DEI/HR veto power—can realign big tech companies with user needs and technological reality.

AI training data is becoming a new revenue stream (TAC 2.0), but its value model is unsettled.

Deals like Google–Reddit and Stack Overflow suggest high-margin licensing income for platforms with rich corpora, yet the hosts stress uncertainty over whether these will be ongoing ‘traffic acquisition’–style payments or more episodic, Netflix-like content buys whose value decays as data ages.

AI will rapidly erode level-one customer support and pressure call centers and SaaS vendors.

Klarna’s AI assistant reducing resolution times, costs, and repeat inquiries shows that frontline support roles are highly automatable, threatening firms like Teleperformance and tools like Zendesk while opening opportunities for leaner, more automated startups.

Eliminating or minimizing HR/DEI functions can preserve speed and meritocracy in startups.

Chamath outlines his practice of having no internal HR in majority-controlled companies—outsourcing serious issues to external employment counsel, letting teams design benefits and run hiring, and rigorously managing out the bottom 5–10% of performers—to avoid internal ‘commissar’ politics and bureaucracy.

WORDS WORTH SAVING

5 quotes

“Every meeting above a certain size [at Google] has a DEI person in it… they’re the commissars, quietly taking notes.”

David Sacks

“There is no company where I have majority control where I have an HR department.”

Chamath Palihapitiya

“The real threat to Google is… are they in a position to maintain their search monopoly under the threat of AI, and are they organized to compete effectively?”

David Friedberg

“This TAC 2.0 thing is amazing… if you’re an entrepreneur building a site with unique data, you’ll be able to license and sell that as an incremental revenue stream.”

Chamath Palihapitiya

“Klarna replaced 700 people and saved $40 million, but Teleperformance lost $1.7 billion of market cap when that tweet went out.”

Chamath Palihapitiya

Google Gemini’s DEI-driven image generation failures and leadership accountabilityThe role and power of DEI, HR, and “Responsible AI” teams inside big techAI training data deals: Google–Reddit, Stack Overflow, and OpenAI’s content licensingKlarna’s AI customer support agent and broader impact on labor and SaaSReddit’s IPO S‑1: user growth, monetization, and long-term valuation questionsThe economics and future of data/content licensing for LLM trainingApple’s cancellation of its car project (Project Titan) and pivot toward AI

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome