@juanhenaoparra: These are the accounts I read the most about AI. 1. @bcherny: leads Claude Code at Anthropic. Shares in first person ho…

X AI KOLs Timeline News

Summary

A curated X thread listing the 10 accounts the author reads most for AI news, covering leaders like Anthropic's Claude Code team, LangChain's CEO, and builders testing new models and agent setups.

🧵 These are the accounts I read the most about AI. 1. @bcherny: leads Claude Code at Anthropic. Shares in first person how he programs with Claude: real prompts, formal verification with Lean/TLA+ and how the team works from the inside. 2. @hwchase17: CEO of LangChain. Thinks out loud about harnesses, memory, async subagents, evals and model routing, with numbers from his own agents in production. 3. @LangChain: updates on LangSmith and Deep Agents, and cases of companies already running agents in production (evals, traces, long context). 4. @rauchg: CEO of Vercel. Data on how agents are already deploying and buying infrastructure, and opinions on CLIs, Rust and how to design software for agents. 5. @amasad: CEO of Replit. Product vision on agents that build apps and businesses, security in desktop apps with AI and the impact of AI on the economy. 6. @santtiagom_: in Spanish, from MercadoLibre. Explains agent concepts (durability, prompt injection, permissions) and shares guides and papers worth saving. 7. @MatthewBerman: tests new models the day they come out and shows how they perform on real tasks. Useful for getting a quick idea before trying them yourself. 8. @0xMovez: threads and long articles on setups with Claude: skills, MCPs, reverse engineering, motion design. Very enthusiastic tone, but the resources he shares are concrete. 9. @kirillk_web3: daily AI news and open source repos moving on GitHub. Good for not missing what's happening. 10. @ParamSiddh: AI engineer with short opinions on vibe coding and which model to use for each thing (planning, executing, reviewing).
Original Article
View Cached Full Text

Cached at: 10/16/26, 06:46 PM

🧵 These are the accounts I read the most about AI.

  1. @bcherny: leads Claude Code at Anthropic. Shares in first person how he programs with Claude: real prompts, formal verification with Lean/TLA+ and how the team works from the inside.

  2. @hwchase17: CEO of LangChain. Thinks out loud about harnesses, memory, async subagents, evals and model routing, with numbers from his own agents in production.

  3. @LangChain: updates on LangSmith and Deep Agents, and cases of companies already running agents in production (evals, traces, long context).

  4. @rauchg: CEO of Vercel. Data on how agents are already deploying and buying infrastructure, and opinions on CLIs, Rust and how to design software for agents.

  5. @amasad: CEO of Replit. Product vision on agents that build apps and businesses, security in desktop apps with AI and the impact of AI on the economy.

  6. @santtiagom_: in Spanish, from MercadoLibre. Explains agent concepts (durability, prompt injection, permissions) and shares guides and papers worth saving.

  7. @MatthewBerman: tests new models the day they come out and shows how they perform on real tasks. Useful for getting a quick idea before trying them yourself.

  8. @0xMovez: threads and long articles on setups with Claude: skills, MCPs, reverse engineering, motion design. Very enthusiastic tone, but the resources he shares are concrete.

  9. @kirillk_web3: daily AI news and open source repos moving on GitHub. Good for not missing what’s happening.

  10. @ParamSiddh: AI engineer with short opinions on vibe coding and which model to use for each thing (planning, executing, reviewing).

Similar Articles

@dengdry: If you're on X to learn about technology and want to stay updated on the latest AI developments, here are the accounts you must follow: Anthropic @karpathy: a must-follow in the AI community, recently joined Anthropic @bcherny: creator of Claude Code, often shares practical tips @…

X AI KOLs Timeline

Recommends key X accounts in the AI field such as Anthropic, OpenAI, Google AI, and xAI, suggesting directly following frontline researchers and developers to get the latest AI updates.