Boris Cherny

OpenAI is quietly testing a paid feature that allows ChatGPT Plus users to immediately reset their weekly usage quota via an "$8 to reset" button. This early experiment, which was not formally announced, enables the removal of the typical waiting period associated with depleted AI usage caps. The company stated it is exploring ways for some users who hit quotas to buy additional usage.

Boris Cherny, creator of Anthropic's Claude Code, advises users to avoid micromanaging AI by providing overly specific instructions. Instead, he recommends focusing on outcomes and allowing modern models to determine the execution path. Andrew Ng of Google Brain previously advocated a similar approach, termed 'lazy prompting,' as models evolve to handle higher-level tasks independently.

Boris Cherny, creator of Claude Code, argues companies should measure AI success by human hours saved rather than token usage. He outlines a framework for tracking AI returns in enterprise workflows and highlights the shift from token-centric metrics as businesses reassess AI costs. Executives from Coinbase, Vercel, JPMorgan, and OpenAI have also discussed strategies for optimizing AI spending and value.

Boris Cherny, creator of Claude Code, outlined five future job archetypes: Prototyper, Builder, Sweeper, Grower, and Maintainer. He suggests these roles may replace traditional domain-specific roles as AI reshapes work, with teams balancing these archetypes based on product maturity. Cherny and others discussed AI's potential to assist across these roles and the need for worker flexibility.

Anthropic's Boris Cherny argues companies should prioritize AI's return on investment (ROI) while allowing employees to experiment with AI tools. He emphasizes that controlled experimentation fosters innovation and that cost management should occur after identifying viable use cases.

Loop engineering is gaining prominence in AI as a replacement for prompt engineering, with figures like Claude Code creator Boris Cherny and OpenClaw creator Peter Steinberger advocating for automated systems that reduce manual prompting. Loops enable AI agents to perform recurring tasks autonomously, such as coding and repository maintenance, by structuring workflows with components like automations, worktrees, and sub-agents.

Boris Cherny, head of Claude Code at Anthropic, states he hasn’t written code manually in eight months, relying entirely on Claude Code. He notes that major clients like Salesforce, NASA, and Y Combinator startups are adopting similar AI-driven workflows, comparing the shift to Gutenberg’s printing press in its potential to democratize software creation.

Anthropic's CFO Boris Cherny discusses Claude Code's $2.5B annualized revenue and advises comparing AI costs to engineering labor rather than traditional software subscriptions. He highlights efficiency gains, such as a developer rewriting a codebase in six days versus a year of manual work, and emphasizes internal pilot testing to measure AI adoption impact.

Anthropic held its Code with Claude developer event in London, showcasing how AI-powered coding tools have dramatically transformed software development. Nearly half of attendees reported shipping pull requests entirely written by Claude without reading the code, demonstrating how quickly LLM-based tools have become normalized in the industry.

OpenClaw users face new restrictions as Anthropic, an AI lab, imposes paid access to its Claude AI for third-party tools. Anthropic aims to reduce system strain and generate profit, with Boris Cherny acknowledging subscription models were not designed for high-usage external applications.

Anthropic's AI model Claude has faced backlash from power users who claim it has regressed in performance, with speculation about deliberate scaling back to prioritize resources for its new Mythos model. Anthropic denies compute constraints or Mythos-related changes caused the issues, attributing the shift to adjustable reasoning settings in Claude Code.