This Week in Startups

What’s Going On With Google Gemini Deep Research? | E2072

194 snips
Jan 16, 2025
Aarush Selvan, Lead PM at Google Gemini Deep Research, dives into the revolutionary advancements in AI, specifically explaining their cutting-edge Deep Research technology that personalizes research experiences. They discuss the challenges of AI reliability, including the 'hallucinations' phenomenon in generated content. Aarush also shares insights on Google's AI business model, the integration of memory in AI systems, and the implications of these innovations for user research. Plus, hear about TikTok's struggles in the current regulatory landscape.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Deep Research Computational Cost

  • Deep Research is computationally expensive, impacting Google's servers.
  • They aim to reduce compute costs with newer models like Gemini 2.0 Flash.
INSIGHT

Deep Research Business Model

  • Google's traditional business model relies on ads, while Deep Research uses a subscription model.
  • They aim to balance compute costs with value for stakeholders, offering source transparency.
ANECDOTE

NotebookLM for Case Study Analysis

  • Jason used NotebookLM to analyze McDonald's case studies and Ray Kroc's biography.
  • This allowed querying a specific dataset, similar to an automated venture capitalist.
Get the Snipd Podcast app to discover more snips from this episode
Get the app