Claude Code prompts a grayed‑out commit button in Emacs, reducing a multi‑step workflow to a single keystroke.
CLAUDE CODE DISRUPTS EMACS WORKFLOWS, AI OUTPERFORMS HUMAN CODERS
*Claude Code entered the scene in September 2025 and quickly eclipsed traditional Emacs tooling. Its ability to automate git commits and solve thousands of research problems in hours signals a shift from rule‑based editors to stochastic AI assistants.*
The Emacs community is confronting a new reality. A year after the author first tried Claude Code in September 2025, the AI assistant has become an order of magnitude more capable, automating tasks that once required deliberate keystrokes. The shift is not a gradual upgrade; it is a wholesale replacement of the manual workflow that defined Emacs culture. At the same time, OpenAI’s latest model demonstrated the power of brute‑force AI research, processing 8,000 problems in three hours and surfacing 372 promising leads. These twin developments—AI‑driven coding assistance and massive problem‑solving capacity—challenge the long‑held belief that human‑crafted LISP scripts are the pinnacle of programmer efficiency. The result is a rapid re‑evaluation of what it means to code, collaborate, and claim authorship in a world where a “god machine” can draft commits in seconds.
Claude Code's rapid rise since September 2025
The author began using Claude Code in September 2025. Within a week the tool “blew my socks off,” and a year later it is described as an order of magnitude more capable. Claude Code now handles the entire coding process, from generating snippets to prompting commit actions. The shift is stark: where Emacs users once relied on manual keystrokes, Claude Code anticipates intent and offers ready‑made prompts. The author’s experience mirrors a broader trend where developers trade hand‑crafted edits for AI‑driven suggestions, reducing the friction of repetitive tasks.
The Mathocalypse: OpenAI's three‑hour blitz
Scott Aaronson labels the latest OpenAI model’s feat as “The Mathocalypse.” In three hours the model carpet‑bombed 8,000 research problems, producing 372 likely hits that could be career‑making if solved by a human. The sheer scale underscores how stochastic AI can outpace traditional, rule‑based problem solving. The result is a flood of plausible solutions that reshape expectations for academic and industrial research pipelines.
Emacs, LISP, and the AI takeover
Emacs has long championed LISP’s expressive power against C’s execution speed. The article recalls RMS’s early circuit‑analyzer work, which favored human‑like diagnostics over equation solving. LISP’s “car” and “cdr” symbols remain cryptic, while C’s oddities like “creat” are at least word‑based. Today, the advantages Emacs users claimed—editing speed, customization, LISP expressiveness—vanish the moment an AI like Claude Code takes over. The editor becomes a conduit for AI prompts rather than a manual craft tool.
Will AI co‑author credits become the norm in code commits?
Claude Code now prompts a grayed‑out “commit” after making a discrete chunk of changes, allowing a quick tab‑return to finalize. The system automatically adds Anthropic as a co‑author in the git history. This raises a practical question: should human developers claim credit, or accept AI attribution as standard? The industry is watching as AI‑generated contributions blur the line between tool and collaborator, potentially reshaping open‑source licensing and reputation systems.
Frequently asked questions
When did the author start using Claude Code and how did it perform?
The author began using Claude Code in September 2025; it impressed him immediately and is now described as an order of magnitude more capable.
How many research problems did OpenAI's latest model process and what were the outcomes?
OpenAI's model processed 8,000 research problems in three hours, yielding 372 likely hits that could be career‑making if solved by a human.
How is Claude Code changing the git commit workflow in Emacs?
Claude Code prompts a grayed‑out "commit" after making a discrete chunk of changes, allowing a quick tab‑return, and adds Anthropic as a co‑author in the git history.
If Claude Code can draft commits, solve research problems, and outpace seasoned Emacs users, the next frontier is institutional acceptance. Developers must decide whether to share credit with their AI partners or to reassert human ownership. As AI co‑author tags appear in git histories, the software ecosystem will need new norms for attribution, liability, and trust. The coming months will reveal whether the community embraces AI as a collaborator or fights to preserve the artisanal roots of Emacs.
Sources: Hacker News – Emacs in the Age of Magic (https://commandlinesystems.com/magic.txt)