Claude Code Founder: 3 Principles Every Team Member Follows
▼ Summary
– Boris Cherny’s core principle is to automate work by having Claude, Anthropic’s AI, perform tasks instead of people.
– He advocates for “underfunding things a little bit” to force teams to rely on and integrate AI tools like Claude.
– Cherny advises starting by giving engineers ample AI tokens (usage budget) to experiment, and only optimizing for cost after a successful project is built.
– The high token costs of AI usage are becoming a significant financial concern, with some CFOs experiencing sticker shock.
– Speed is a critical principle, and using Claude to automate tasks is a primary method for accelerating development and competing in the market.
Boris Cherny, a founder at Anthropic, champions a straightforward set of principles for his team, with the AI assistant Claude playing a central role in their workflow. He emphasizes that automating work through Claude is a powerful lever for productivity, a pattern he observes consistently. During a recent podcast appearance, he shared three core tenets that guide his team’s approach to building and innovation.
The first principle is elegantly simple: if a task can be delegated to Claude, it should be. The mindset shifts from “how do I complete this?” to “how can Claude handle this?” This philosophy actively encourages team members to identify and automate repetitive processes, freeing up human creativity for more complex problems. Cherny believes this is fundamental to scaling effectively.
Secondly, Cherny advocates for strategically underfunding projects in their initial stages. This deliberate constraint forces engineers to lean heavily on AI tools from the start, a process he calls “Claude-ifying” their work. The idea is to build with automation as a core dependency, not an afterthought. However, he cautions against premature cost-cutting. His advice to technical leaders is to begin by providing engineers with abundant resources, or “tokens,” for experimentation. The goal is to foster innovation without early restrictions. Optimization for cost should come later, only after a valuable prototype or feature has been proven. This reflects a growing industry conversation, as some companies experience surprise at the operational expenses linked to widespread AI adoption among developers.
The final guiding principle is a relentless focus on speed. In a competitive landscape where rivals can release major updates nearly simultaneously, moving quickly is a critical advantage. Cherny notes that this was essential from the beginning when a small team’s primary edge was velocity. Today, Claude remains instrumental in maintaining that pace. By offloading substantial portions of development and testing to the AI, teams can iterate and ship products dramatically faster. This approach was demonstrated when Anthropic used Claude to help build a new agent in just ten days. For Cherny, the directive to move faster and the tool to enable it are perfectly aligned, using Claude for more tasks directly fuels accelerated progress.
(Source: Business Insider)
