This episode reveals how Ridges is building autonomous AI software engineers on Bittensor, leveraging a unique economic model to create a sustainable, open-source alternative poised to undercut cash-burning centralized competitors like Cursor and Claude.
Introduction to Ridges: The AI Software Engineer
- The Problem: Current AI coding tools, like Cursor or Claude, act as an "Iron Man suit" for human engineers. They augment the developer's capabilities but still require a skilled human operator who understands code.
- The Ridges Solution: Ridges aims to create a "robot army" of autonomous AI engineers. A non-technical user can describe a task in plain English, and the AI engineer will attempt to complete it, review its own work, and deliver a final product without human coding intervention.
- Shaq frames the ultimate vision as an HR tool for managing AI engineers rather than a development tool for human engineers. As he puts it, "The final form of this is basically like an HR tool where you're just trying to like managing this AI to do it for you as like a regular person would manage an engineer."
The Economic Advantage: A Sustainable Model for AI Development
- The conversation highlights Ridges' significant cost advantage over competitors. While a human engineer costs $10-20k per month and a tool like Claude costs around $3,500 per month, Ridges is targeting a price point of $1,000 per month for an AI engineer.
- Thick Agent, Thin Model: Unlike competitors who rely on massive, expensive AI models (a "thick model") with a simple interface layer, Ridges focuses on building a highly sophisticated "thick agent layer" that can extract maximum performance from smaller, cheaper, open-source models like DeepSeek V3. This makes Ridges less dependent on costly, proprietary models.
- Demonstrated Adaptability: Shaq shares a key example of this strategy's success. When miners were forced to switch from a superior model to a less capable one, their performance score dropped from 45% to 30%. Within two days, by improving only the agent layer, they recovered to the original 45% score using the cheaper model.
- The Bittensor Composability Effect: Host Mark Jeffrey points out that Ridges compounds its savings by using other subsidized Bittensor subnets like Shoots and Targon for model inference. This creates a cascading cost-saving effect throughout the AI supply chain, with the Bittensor network's TAO emissions subsidizing each step.
Driving Innovation Through Open Competition
- Ridges operates on a "winner-take-all" competitive model where hundreds of developers (miners) compete to build the best AI agent. This continuous, decentralized R&D process is a core part of its strategy.
- Fiat On-Ramp for AI Talent: Ridges recently launched a dashboard that allows AI developers to compete without any crypto knowledge. They can sign up with a Google account, submit their agent, and if they win, Ridges converts the TAO emissions into fiat and sends it directly to their bank account.
- Strategic Implication: This move is critical for attracting top-tier AI talent who may be averse to or unfamiliar with crypto, massively expanding the potential talent pool. The top agent on the platform can earn nearly $40,000 in a 24-hour period.
- Emergent Complexity: The competition is fostering sophisticated strategies. Shaq describes one miner who created a "team of like agents that are crawling through the codebase," communicating their findings to a central "home base" agent to solve problems more efficiently. This demonstrates a level of innovation that would be difficult to replicate in a closed, centralized team.
Go-to-Market Strategy and Enterprise Adoption
- Ridges is nearing its first product launch, an IDE (Integrated Development Environment)—a software application that provides comprehensive facilities to computer programmers for software development. This product will be powered by the winning agents on the network.
- Unsustainable Competitors: Shaq notes that competitors like Cursor are reportedly losing five dollars for every dollar of revenue earned, leading them to quietly degrade their service over time by switching to cheaper, less effective models.
- Ridges' Positive Unit Economics: In stark contrast, Ridges' inference cost is approximately 1/250th of its competitors. This allows Ridges to offer a product that continuously improves as its agents get smarter, all while maintaining a profitable margin from day one.
- Solving the Enterprise Privacy Problem: A major hurdle for enterprise adoption of AI coding tools is the need to send proprietary code to third-party servers. Ridges solves this by being fully open-source. Enterprises can license and run the top-performing agent locally, ensuring their code never leaves their own infrastructure.
Why Build on Bittensor?
- Shaq explains that building on Bittensor was a strategic choice that provides advantages unavailable in the traditional venture-backed startup world.
- Incentivized R&D at Scale: The Bittensor network effectively subsidizes Ridges' research and development by paying the competing miners. Shaq estimates this subsidy amounts to $15 million per year—a sum that would be unsustainable for a traditional startup to pay out in cash for R&D.
- Battle-Tested Infrastructure: Bittensor provides a ready-made, battle-tested platform for running secure, global competitions, handling everything from cheat prevention to payment distribution.
- A New Paradigm: This model allows Ridges to focus on packaging the winning technology into a product, while the core innovation is driven by a decentralized, global competition.
Market Outlook and Ecosystem Favorites
- The discussion touches on the broader Bittensor ecosystem and the challenges of subnet tokenomics. While acknowledging the current sell pressure from high emissions, Shaq is optimistic that breakout subnets providing real-world, non-financial utility will attract significant, non-speculative capital into the ecosystem.
- Favorite Subnets: Besides Shoots and Targon, which power Ridges, Shaq highlights several other promising projects:
- Red AI & Score: Interesting business models.
- Bitcast: A decentralized video marketing platform that Ridges uses for cost-effective content creation.
- BitMind: A deepfake detection subnet that Shaq believes is highly underrated and will become critical infrastructure as AI-generated content proliferates.
- Taoshi: A subnet focused on financial models and trading algorithms.
Conclusion
This episode demonstrates that Ridges is not just building another AI tool; it is pioneering a new, economically sustainable model for AI development. By leveraging Bittensor's decentralized competition and composable infrastructure, Ridges is creating a product with superior unit economics and a clear path to enterprise adoption, posing a legitimate threat to centralized incumbents.
