Perplexity AI Google Launches Gemini 3.1 and YouTube AI
18 snips
Feb 20, 2026 Discussion of Gemini 3.1 Pro's rapid release cadence and why incremental fine-tuned updates matter. Conversation about real-world leaderboards and blind human comparisons versus vendor benchmarks. Overview of Gemini powering YouTube TV with live TV assistants and practical TV-based AI queries. Highlights of YouTube's broader AI rollouts like upscaling, comment summarizers, and creator tools.
AI Snips
Chapters
Transcript
Episode notes
Rapid Incremental Model Iteration
- Google is iterating quickly on Gemini with 3.1 as a focused, incremental upgrade rather than a full overhaul.
- Jaeden Schafer highlights that small, composable improvements often roll into future major releases.
Use Incremental Releases To Adopt Features
- Expect incremental features from model updates to persist into future major releases.
- Use 3.x releases to test and adopt practical tools before larger model changes arrive.
Trust Blind Leaderboards Over Vendor Screenshots
- Public benchmarks can be gamed, so blind human leaderboards carry more credibility for real-world performance.
- Jaeden says he trusts blind comparisons and real-world leaderboards more than vendor-selected screenshots.