Cameron Fairchild: Your Next Customer Is an AI Agent, Bittensor, Ridges, Gamma Tokens | Ep. 102

In this episode, Cameron Fairchild, CTO of Latent Holdings, joins Grant at Exploit Summit in Montréal to discuss building products for an economy where AI agents discover services, delegate work, and make payments. He discusses Ridges’ shift toward specialist coding agents, Cacheon’s inference optimizations, and why real-world usage matters more than performance on a public benchmark. Cameron also details the distribution challenge of marketing to agents, the distinction between subnet revenue and profit, the proposed role of gamma tokens as service credits, and the governance questions that emerge as Bittensor develops into a broader economy. *In this episode, you’ll learn* - Why Ridges is moving beyond public benchmark scores toward realistic workloads, specialist agents, and evaluation tied to the cost and value of completed tasks - How a mixture of specialist agents could handle linting, database work, and test generation, and why routing tasks to the right agent becomes its own optimization problem - How Cacheon’s kernel optimizations can accelerate inference without access to a customer’s proprietary model weights, and why energy costs matter as inference becomes more efficient - How agents could discover Ridges through API documentation and pay through x402, allowing services to gain customers without owning the interface those customers use - How Cameron uses coding agents and layered agent reviews to extend engineering capacity and improve coverage during Bittensor development - Why revenue alone can hide negative gross margins, how protocol subsidies complicate measuring subnet performance, and why Cameron distinguishes this risk from an accusation against current subnets - The proposed conversion of alpha into gamma service credits, the comparison with future compute purchases, and the redemption risks that affect what those credits are worth - Why Cameron sees circular spending of TAO and alpha as important to Bittensor’s economy, allowing participants t

Watch the video on Ventura Labs (YouTube) ↗

Taosis indexes announcements, articles, posts and releases about Bittensor subnets from their own channels and keeps a record of each. The text above is the source’s; the figures are read from the chain by Taosis.