为什么现在大家评价 LLM，越来越看重代码和 agent 能力，而不是聊天能力？我的两点观察：1. 可能是我身处计算机领域的“领域视角”带来的偏差——圈子里天天聊 Claude Code、Codex，感觉全世界都在用 AI 写代码2. 但更本质的是：人们不再满足于模型"会聊天"，而是希望它能把活干完。聊天是过程，交付才是结果说实话，普通人的日常需求，豆包这类产品早就够用了。卷代码和 agent，卷的是生产力的天花板。#LLM #AI编程 #Agent #ClaudeCode #AI
Related
📰 Australian STEM Videogame Challenge Canceled Due To AIPC Gamer reports: The Australian STEM Videogame Challenge, a "fr...
📰 Australian STEM Videogame Challenge Canceled Due To AIPC Gamer reports: The Australian STEM Videogame Challenge, a "free national videogame development competition for students,"...
Muhammad Zulali ha collegato Claude Code e Codex alla centralina della sua Jeep Cherokee del 1994 con un Arduino Nano da...
Muhammad Zulali ha collegato Claude Code e Codex alla centralina della sua Jeep Cherokee del 1994 con un Arduino Nano da circa 3 dollari. Gli agenti hanno letto codici e parametri,...
#FediversePOSTs on Fediverse ( 2026-11-14 )https://mnmm.top/oMz[Tips are appreciated.]https://mnmm.top/tipマクドナルドでの朝食からスタ...
#FediversePOSTs on Fediverse ( 2026-11-14 )https://mnmm.top/oMz[Tips are appreciated.]https://mnmm.top/tipマクドナルドでの朝食からスタートし、相模大野ステーションスクエアのツリーや歴史パネル展、書店やスーパーを巡ってお買い物を満喫。しかし、前日のお酒の悪...