Bozhidar Batsov, a 20-year Emacs maintainer and author of some of the ecosystem's most widely used packages, published a detailed analysis in March 2026 examining how the AI coding revolution intersects with Emacs and Vim. He maps out specific risks and opportunities facing both editors rather than a simple verdict.
On the risk side, he identifies three structural pressures: the gravitational pull of AI-native editors like Cursor and Windsurf toward VS Code's ecosystem, the diminishing case for mechanical editing speed when AI is generating most code, and a corporate backing asymmetry that leaves volunteer-driven projects like Emacs and Neovim unable to match the engineering resources Microsoft and well-funded startups can deploy. In an extreme scenario, Batsov notes that full programming automation would render the entire coding editor category obsolete — though he considers this unlikely in the near term.
The more compelling thread in Batsov's essay concerns terminal-native composability. Tools like <a href="/news/2026-03-14-1m-token-context-window-generally-available-claude-opus-4-6-sonnet-4-6">Claude Code</a>, Aider, and Google's gemini-cli work naturally alongside Emacs and Vim without requiring dedicated editor integrations, meaning users can already capture much of AI's productivity benefit without abandoning their preferred workflow. Batsov also highlights the Agent Client Protocol (ACP) — framed as an LSP equivalent for AI agents — as a direct path to agent integration in Emacs. ACP decouples agents from editors so any compatible agent works with any compatible editor, giving Emacs access to the broader agent ecosystem without bespoke glue code for each tool. This is directly evidenced by agent-shell, a native Emacs mode built by developer xenodium specifically motivated by ACP's arrival, which now supports Claude Code, GitHub Copilot, Cline, Kiro CLI, and others.
The Hacker News discussion that followed Batsov's essay shows how these dynamics play out among working developers. Xenodium confirmed that the lack of an agent-agnostic integration path was the primary barrier to adopting popular agents throughout most of 2025, and that ACP's emergence directly motivated agent-shell's development. Another commenter described keeping Emacs as a dedicated human-driven environment while using Claude Code and gemini-cli separately for several hours each week — a workflow that treats terminal-native AI tools as composable utilities rather than editor features. Batsov also notes that AI substantially lowers the barrier to Elisp and Lua configuration, which has historically been the single biggest obstacle to Emacs and Vim adoption. Whether that democratization offsets the pull of <a href="/news/2026-03-14-nyt-ai-coding-assistants-end-of-programming-jobs">purpose-built AI editors</a> is the question his analysis leaves deliberately open.