![](\"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/06\/55323004951_b431494339_6k.jpg?w=2048\")
<\/p>\n<\/p>\n
The man who built the tool that\u2019s rewriting how software gets made hasn\u2019t touched a keyboard to write code in the better part of a year.<\/p>\n
Boris Cherny, the head of Claude Code at Anthropic, dropped the detail almost in passing during a fireside chat at the Fortune<\/em> Brainstorm Tech conference. \u201cI haven\u2019t written a line of code by hand in, I think, eight months now,\u201d he told Fortune<\/em> AI editor Jeremy Kahn: \u201cClaude Code, 100% written by Claude Code,\u201d he said. He added that Anthropic\u2019s biggest enterprise customers \u2014 Salesforce, NASA, Y Combinator startups \u2014 are trending in the same direction.<\/p>\n This is the reality taking shape inside what may be the closest thing the tech industry has to a fully agentic organization. And if Cherny is right about where it\u2019s heading, the implications stretch far beyond Silicon Valley.<\/p>\n Cherny reached for a 500-year-old metaphor to explain what\u2019s happening. Before Gutenberg\u2019s printing press in the 1440s, European literacy hovered around 10%. Reading and writing were professional skills, the province of scribes employed by lords and kings. The press didn\u2019t just make books cheaper \u2014 it reduced the cost by 100x and triggered an explosion of published literature that exceeded the previous thousand years in just five decades.<\/p>\n \u201cWhat happened was just insane, and something that no one could have expected,\u201d he said. It took a couple hundred years, he noted, for education systems to evolve with the explosion in literature, \u201cbut global literacy went up.\u201d The Renaissance, the Reformation and the Industrial Revolution wouldn\u2019t have happened without this unlock, he argued.<\/p>\n \u201cClaude Code is democratizing people\u2019s ability to write software,\u201d he said. \u201cWhat does it mean if, you know, in the past there were 50 million people in the world that could code, and now everyone in this room can code?\u201d<\/p>\n The analogy is imperfect \u2014 he acknowledged, noting that the printing press also toppled ideologies and unleashed devastating religious wars \u2014 but the underlying metaphor of software as the literacy of the digital economy holds. The people who could write it shaped the institutions, products, and power structures of the last 40 years. What happens when that barrier falls?<\/p>\n The morning of the talk, Cherny said, he had been managing a few hundred AI agents. Some days, he said, it\u2019s thousands or tens of thousands. For most executives still wrestling with how to get a single chatbot deployment to work reliably, the number sounds like science fiction. But Cherny described a structure that\u2019s becoming the new normal inside Anthropic: Claude Code doesn\u2019t just respond to human prompts \u2014 it orchestrates sub-agents that are themselves Claude instances. Human prompting, he noted, is increasingly the exception. \u201cIf you look at most Claude Code sessions, it\u2019s actually another Claude that does the prompting.\u201d<\/p>\n Anthropic recently launched what it calls dynamic workflows, designed to scale this architecture further \u2014 enabling massive parallel tasks such as full codebase migrations or iterative security profiling that would previously have required large engineering teams and months of work. Case in point: developer Jared Sumner recently rewrote the Bun JavaScript runtime from Zig to Rust using Opus 4.8 and dynamic workflows. Cherny claimed that the estimated timeline with a human engineering team would be roughly a year, but Sumner\u2019s actual time was six days.<\/p>\n One of the more practically useful insights Cherny shared is what he called a bottleneck migration problem \u2014 and it\u2019s something any executive deploying AI at scale will eventually run into. Automate one stage of a process, and you don\u2019t eliminate friction; you move it. At Anthropic, the sequence has played out like this:<\/p>\n \u201cFind the bottleneck, solve the bottleneck,\u201d Cherny said. \u201cAnd anytime you have to do a task, build a skill that will solve similar tasks in the future.\u201d<\/p>\n For enterprise buyers getting sticker shock from token costs, Cherny offered a simple but significant reframing: stop comparing Claude Code to your $20-a-month coding assistant. Compare it to what an engineer would have cost to do the same work.<\/p>\n \u201cThat\u2019s the benchmark,\u201d he said. The Bun rewrite is his Exhibit A. He also recommended internal \u201cshootouts\u201d \u2014 give one team Claude Code, withhold it from another, and measure delivery speed, security, and polish. The data, he argues, builds the ROI case faster than any vendor pitch.<\/p>\n Perhaps the most striking moment of the conversation came when Cherny was asked about recursive self-improvement \u2014 Anthropic\u2019s own recent blog post flagged the company\u2019s code output has grown roughly 8x compared to the 2021-2025 baseline, largely because Claude is writing Claude. He described Claude Code as potentially \u201cthe first product that actually just takes off\u201d because it\u2019s fully writing, reviewing, and security-scanning itself, and is beginning to generate its own feature ideas by scanning GitHub issues, Twitter, and Slack.<\/p>\n \u201cMany mornings I wake up, and Claude already has pull requests that it came up with, verified end to end, it has screenshots for me,\u201d he said.<\/p>\n When asked whether he was worried about how fast this is moving, he answered without hesitation: \u201cYes \u2026 It\u2019s one of the big risks for AI.\u201d<\/p>\n Still, Anthropic\u2019s blog post cautioned that 8x productivity jump was \u201calmost certainly an overstatement\u201d because measuring lines of code rewards volume, not quality. To return to the Jared Sumner example, that compelling anecdote should also be treated with extreme caution as business strategy, as Sumner is an elite developer working on an open-source project he created and knows more intimately than anyone else on the planet, making him the best possible human in the loop to ensure that the vibe coding was successful.<\/p>\n Cherny noted that even Anthropic\u2019s expensive \u201cteam of Claudes\u201d code review approach catches \u201cpretty much every bug.\u201d To paraphrase a favorite phrase of Claude\u2019s, pretty much is doing a lot of work in that sentence.<\/p>\n Perhaps the most underexplored thread in Cherny\u2019s talk came from an audience member, not from the stage: when employees stop asking colleagues where the codebase is, when new engineers never need to meet their manager to get unstuck, what organizational tissue quietly dies?<\/p>\n Cherny\u2019s answer was honest and revealing: \u201cThis is something I\u2019ve actually heard from new engineers on the team, that because they\u2019re talking to Claude so much, they don\u2019t get a chance to meet the team as much.\u201d He said Anthropic has started to \u201cconsciously\u201d encourage \u201cpeer programming, so you don\u2019t just pair with Claude, but you also sit there, maybe with another engineer on our team.\u201d They do a lot of \u201csocial time,\u201d he said, \u201cbecause in this environment where we\u2019re actually wrong and our guesses are incorrect a lot, you have to feel very safe being wrong.\u201d<\/p>\n For this story,\u00a0<\/em>Fortune\u00a0journalists used generative AI as a research tool. An editor verified the accuracy of the information before publishing.<\/em><\/p>\n<\/div>\n #Claude #Code #hasnt #written #line #code #hand #months<\/p>\n","protected":false},"excerpt":{"rendered":" The man who built the tool that\u2019s rewriting how software gets made hasn\u2019t touched a keyboard to write code in the better part of a year. Boris Cherny, the head… <\/p>\n","protected":false},"author":1,"featured_media":6819,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[361,536,3849,9215,6853,997,1911,3850,9217,2156,1854,94,455,9216],"class_list":["post-6818","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-finance-news","tag-ai-agents","tag-anthropic","tag-brainstorm-tech","tag-chief-financial-officer-cfo","tag-claude","tag-code","tag-engineering","tag-fortune-brainstorm","tag-hand","tag-hasnt","tag-line","tag-months","tag-software","tag-written"],"_links":{"self":[{"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/posts\/6818","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6818"}],"version-history":[{"count":0,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/posts\/6818\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=\/wp\/v2\/media\/6819"}],"wp:attachment":[{"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6818"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6818"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.fintechpulse8.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6818"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}The Gutenberg moment<\/h2>\n
Managing hundreds of agents before breakfast<\/h2>\n
The bottleneck migration problem<\/h2>\n
\n
The ROI question every CFO is asking<\/h2>\n