Revenue Search: Inside Bittensor Talking Tao: SN62 Ridges Special AMA
57 snips
Oct 7, 2025 In this engaging discussion, Shaq (Shak), the founder of the Ridges project, dives deep into innovative incentive structures that prioritize genuine user engagement over traditional benchmarks. He reveals the upcoming Ridges V1 as a coding-assistant extension, detailing its user-friendly pricing and architecture. Shaq emphasizes the importance of iteration speed for competitive advantage and discusses strategies to prevent gaming the system. With plans for growth and a focus on organic reach, Ridges aims for robust user engagement while ensuring low subscription costs.
AI Snips
Chapters
Transcript
Episode notes
Pay For Progress, Not Payroll
- Ridges pays only the engineers whose agents demonstrably improve the product, not all staff.
- This creates a rapid iteration engine that accelerates agent improvement beyond traditional firms.
Iteration Speed Beats Static Scores
- Mixing benchmarks (Sweebench + Polyglot) dropped scores from ~88% to ~17–18%, then rebounded to ~41% within weeks.
- Fast iteration speed, not a static lead, is their competitive edge.
Product Signals Decide Rewards
- Ridges will run benchmark evaluation then A/B test top agents live with real users before awarding emissions.
- Emissions go to agents that increase accept rate, reduce fixes, and raise engagement.