AI summary by TwinMind
Is Google Getting Worse? (Update) | Freakonomics Radio
Freakonomics Radio Network
56m 53s
Perplexity AI Introduction
- Steven Dubner introduces new AI search competitor Perplexity
- Episode replays earlier "Is Google Getting Worse" content with new Perplexity interview
- Neva (previous Google competitor) already shut down
Perplexity CEO Interview - Aravind Srinivas
- Company Description: Conversational answer engine providing answers with source citations
- Key Differentiators from ChatGPT:
- Accuracy through real-time internet facts vs hallucination
- Speed: 6-7 seconds vs ChatGPT's "eternity" (close to 10 seconds)
- Always pulls relevant facts from internet for factual grounding
- User Base: 10 million monthly active users, 70-100k paying $20/month premium
- Market Competition: AI chatbot/answer bot sector won't be monopoly
- OpenAI, Meta, Microsoft, Apple all competing
- Google "too lethargic" - could have built ChatGPT in 2020-21
- Business Model: Free version + premium paid service
- Users realize time/data value being "cheated" by free ad-supported models
- Worth "few lattes a month" for better service
- Acquisition Speculation: Prefers independence but hypothetically Apple would be "fantastic fit"
- Apple's user access, native integration across devices
- Siri could finally ground itself in real information
- Apple is "one company Google fears"
- Privacy Approach:
- Transparent data usage controls
- Can turn off data usage while retaining history (vs ChatGPT)
- Every query private by default unless user chooses to share
- Better than Google's default indexing of Bard chats
Google's Early History and Rise
- Pre-Google Era: Alta Vista, Fast (Norway), Dogpile meta-search
- Marissa Mayer Background: 20th Google employee, first female engineer, 13 years at Google
- PageRank Innovation: Algorithm ranking pages by incoming links (like academic citations)
- Mission: "Organize world's information, make it universally accessible and useful"
- Early Skepticism: "Why does world need another search engine?"
- Fan Mail: People sent cash/checks saying "search engine so good, not making money, wanted to pay"
Google's Monetization Strategy
- Initial Funding: $25M VC investment in 1999, $75M valuation
- Revenue Attempts: Licensing technology, hardware for internal search
- Advertising Contradiction: Original paper warned "advertising income often provides incentive for poor quality search results"
- AdTech Innovation: Position auctions - bidding system for ad placement
- Auction Mechanics: Based on bid amount + ad quality + relevance + context
- Revenue Dominance: 81% of Google revenue from search advertising
- Financial Analysis: Subscription model ($20/year) vs ads ($52+ per user annually)
- Q4 retail CPMs "well north of $100" vs original $10 estimates
- Ad model 2.5x more lucrative than direct consumer payment
The 8-Year Accidental Experiment
- Setup: 2000 experiment - 99% users get ads, 1% no ads
- Forgotten Experiment: Ran until 2008 when advertisers complained
- Results: 3% more searches from users WITH ads than without
- Conclusion: Ads actually improved search experience
- Reasoning: Advertisers willing to pay indicate higher quality results
- Commercial Intent: Madonna tour tickets example - ads provided actual tickets vs poor organic results
Search Quality and Monopoly Concerns
- Ryan McDevitt Research: Duke economist studying plumber quality
- Yellow Pages Analogy: A-AAA names to appear first alphabetically
- Quality Correlation: Plumbers with many A's were more expensive, lower quality
- Google Application: Same manipulation tactics apply to Google ads
- Complaint Statistics: Google advertising plumbers had 13x more complaints than non-advertisers
- Logic: Firms targeting uninformed, quick-searching consumers can "rip off" customers
- Locksmith Scam: Shady call centers with fake listings, low advertised prices, 3-4x actual charges
Market Dominance and Competition
- Market Share: Google handles ~90% global search activity
- Browser Dominance: Chrome ~67% global market share, Safari ~20%
- Default Search Payments: Google pays Apple reported $18B annually for Safari default
- Competition: Bing (US), Yandex (Russia), Baidu (China), DuckDuckGo (privacy-focused)
- Larry Page 2012: "Competition only click away" - but defaults matter
Yelp vs Google Antitrust Issues
- Jeremy Stoppelman: Yelp CEO/co-founder, started 2004 for local reviews
- Early Partnership: Google initially licensed Yelp content
- Failed Acquisition: Google tried to buy Yelp, then competed directly
- Self-Preferencing: Google prioritized own reviews over Yelp in results
- Senate Testimony 2011: Stoppelman accused Google of monopoly abuse
- Current State: Google reserves "prime real estate" for ads and own content
Web Quality Degradation
- Marissa Mayer Analysis: Web quality decline due to economic incentives
- Scale Change: 30 million pages (2000) to trillion+ URLs today
- Fraud Increase: More misinformation, clicks-for-profit, purchase manipulation
- Ecosystem Problem: Requires multiple actors to solve, not just Google
- Vitali Borker Case: Fraudulent eyeglass seller used customer complaints to boost Google ranking
- SEO Gaming: Constant attempts to manipulate algorithm rankings
Google's Response and Current State
- Liz Reed Interview: VP of Search at Google since 2003
- Ad Revenue Defense: Democratizes access vs premium model for "small subset"
- Usage Scale: Over 1 billion daily users
- Search vs Ads Separation: Claims organic results uninfluenced by advertising
- User Intent Recognition: Different users want different things (speed vs depth)
- Featured Snippets: Inline results to keep users on Google vs sending to external sites
- Quality Vetting: Extensive process for data source accuracy
- Feedback Systems: Users can report errors in Google-generated results
Antitrust and Legal Challenges
- Multiple Cases: DOJ alleging Google "crippling competitive process"
- EU/India Fines: Already fined for antitrust violations internationally
- Location Data Settlement: $400M settlement with 40 states for tracking opted-out users
- Regulatory Shift: From eye-rolls to serious antitrust attention
- Obama Era: Google was #1 most frequent corporate White House visitor
Current Assessment and Future
- Ryan McDevitt Verdict: "Very happy with tradeoff" - few bad links acceptable
- Trade-off Question: Vast Google ecosystem funded by search monopoly
- User Expectations: Constantly rising bar for search quality
- Complex Tasks: Still difficult for multi-step processes like home buying
- AI Competition: ChatGPT, Perplexity, others challenging search paradigm
- Google's Gemini: Company's own AI response to competitive threats
