Google’s search engine “panic,” AI-generated hits, Coinbase considers leaving US market | E1723

Episode Summary

Executive Summary: The episode centers on how generative AI is reshaping search, media, and identity. Jason and Rachel discuss Samsung's reported threat to replace Google with Bing, Google's response via a new personalized search project, and how Google's data moat may help it beat OpenAI and Microsoft. They then explore viral AI voice cloning in music, legal and ethical questions around sampling/copyright, and a demo of AI-generated versions of Jason and Jordan Peterson. The show closes with Coinbase's possible move outside the U.S. and rising pressure around data licensing and AI scraping.

Main Topics: Samsung, Bing, and Google’s search monopoly (Priority: 5/5): The hosts unpack reports that Samsung is considering swapping Google for Bing as the default search engine, and frame it primarily as a negotiation tactic to extract better terms from Google rather than a true competitive threat. Google’s AI search strategy and Project Maggie (Priority: 5/5): They discuss Google’s response to OpenAI and Microsoft, including a new internal search effort codenamed Maggie that aims to deliver more personalized, predictive search using Google’s existing data assets like Gmail, YouTube, and Docs. AI-generated music and viral voice cloning (Priority: 5/5): A fake Drake/The Weeknd track goes viral, prompting a broader discussion of how AI voice models are making synthetic songs increasingly convincing and commercially disruptive. Copyright, fair use, and the future of sampling (Priority: 4/5): Jason argues that AI music will trigger legal battles similar to sampling, with artists and labels seeking payment or control over synthetic versions of their voices and catalogs. AI-generated podcast guests and voice models (Priority: 4/5): The producers demo AI-generated versions of Jason’s voice and Jordan Peterson, showing how well modern voice cloning captures cadence and inflection and opening the door to new show formats. Coinbase and regulatory flight risk (Priority: 3/5): Jason reacts to Brian Armstrong saying Coinbase may relocate if U.S. crypto regulation remains unclear, interpreting it as a serious signal that the company may already be planning a move abroad. AI, data ownership, and platform licensing (Priority: 4/5): The discussion expands to Reddit, Quora, and platform data licensing, with Jason predicting lawsuits and licensing fees will become a major revenue model for AI training data.

Key Arguments: Samsung’s reported Bing consideration is less about switching search providers and more about forcing Google to pay more for default placement. Google can still win the AI race because it owns high-value consumer data (Gmail, YouTube, Docs) that can make search and assistants far more personalized than rivals. The real search contract that matters is Apple’s, not Samsung’s, because iPhone users are wealthier and generate more valuable ad clicks. AI voice cloning is moving from novelty clips to convincing full songs, and eventually full albums, much faster than critics expect. Music rights holders will likely treat synthetic voices like sampled audio: some uses will be licensed, some blocked, and litigation will shape the market. Platforms such as Reddit may demand payment for training data and could even seek injunctions against AI systems that trained on their content without permission. When the hosts’ voices were cloned, the AI captured cadence and inflection well enough to suggest a new category of content production and parody. Brian Armstrong’s public comments about leaving the U.S. likely mean Coinbase is already exploring relocation options. Default settings on smartphones are strategically valuable because users rarely change them, so search distribution is a major economic moat.

Data Points: Samsung-to-Google search deal value: around $3 billion per year - Estimated annual revenue Google gets from Samsung default search placement Apple default-search deal value (2021): around $15 billion - Reported amount Google paid Apple to remain the default search engine on Apple devices Earlier Apple deal value: around $10 billion per year - Previous reported payment Google made to Apple before 2021 Google search market share: 70%–90% - Jason cites this as Google’s dominance across different regions Apple iPhones sold in fiscal 2022: 232 million - Used to illustrate the scale and value of Apple’s user base Apple iPhone sales revenue: around $205 billion - Referenced as annual revenue generated by iPhone sales Average spend per Samsung phone: around $357 - Rough estimate based on Samsung consumer phone sales data Average spend per Apple phone: around $883 - Rough estimate showing Apple users spend far more on devices Samsung vs Apple spend ratio: about 2.5x - Derived from the estimated average spend per new phone LinkedIn users: 875 million - Mentioned during the LinkedIn Jobs ad read AI Drake/The Weeknd viral views on TikTok: more than 8.5 million - BBC-reported view count for the synthetic track AI Drake/The Weeknd Spotify streams: over 250,000 - Reported streaming count before takedown UMG market share: 37.5% - Estimate of Universal Music Group’s share of the music industry Drake re-signing reportedly worth: as much as $400 million - Referenced as context for UMG’s control over Drake’s catalog Model training audio for Jason voice clone: about 30 minutes - Amount of high-quality audio used to train the voice model AI voice demo length: 15 seconds - Snippet of Jason’s AI-generated intro played on the show

Pivotal Quotes: "Google is reportedly in panic mode and working on an entirely new search engine" — Rachel: Introductory framing of the Samsung/Bing/Google report "This is just Samsung having a great negotiator who leaked this information to the New York Times." — Jason: Jason argues the Samsung story is mainly leverage in a negotiation with Google "I am long Google in this fight. I'm also long Microsoft in this fight." — Jason: Jason’s stance that competition will accelerate product improvement rather than destroy Google

Implications: Search defaults, AI voice cloning, and training data rights are becoming core battlegrounds. Expect more personalized AI products, more licensing fights, and faster normalization of synthetic media across music and podcasts.