Can Demis Hassabis save Google?
State of Google: Does It Need “Saving”?
- Some argue Google is technically strong (top-tier LLMs, vast compute, data, product surface) and not in existential trouble.
- Others think it faces an “innovator’s dilemma”: core ad/search model conflicts with AI products that might cannibalize search.
- There is concern that AI competitors (OpenAI, Anthropic, Perplexity, Kagi, etc.) are eroding search dominance, especially for complex queries.
Gemini and LLM Quality
- Several commenters report strong experiences with Gemini Advanced, especially for long-context writing, code help, and up-to-date information.
- Others find Gemini unreliable compared with competitors for programming tasks, even on simple code formatting.
- Some see Gemini as technically powerful but over-censored and untrustworthy on political/geopolitical topics.
- Overall view: Gemini is competitive in some domains (vision, long context) but not consistently “best” across the board.
Monetization, Ads, and Business Model
- Major open question: how to monetize LLMs at the scale and margins of search ads.
- Some think LLM queries will make ad targeting easier; others doubt keyword-style auctions will translate.
- Widespread belief that ad-based web economics are under threat as AI becomes the main interface, potentially stripping out ads and low-quality content.
- Counterpoint: Google still has massive ad surfaces (search, YouTube, Maps, Gmail) and could integrate AI-driven ads and subscriptions.
Leadership, Culture, and Strategy
- Strong criticism of current management: seen as optimizing for short-term ad revenue, letting product quality and culture decline.
- Many see the main problem as leadership, not engineering talent; Google is described as mismanaged and slow, with muddled priorities.
- Comparisons are drawn to past tech giants that failed to respond to platform shifts.
Role of Demis Hassabis and DeepMind
- DeepMind’s work (e.g., game-playing systems, scientific models, Gemini) is framed as Google’s “lifeline.”
- Some think he should lead the whole company; others say CEO duties (sales, org politics) would waste his research talent.
- Skepticism exists about “savior” narratives; many argue one person or “bolted-on AI” cannot fix deep structural and cultural issues.