February 16, 2026
Keyboards vs. chatbots: FIGHT
10 years building vertical software: are we cooked?
Stocks tumble, chatbots crash the party, and the comments yell: panic or chill
TLDR: A veteran founder says AI chat agents gut the “moats” of expensive, hard-to-learn industry software by making training and switching costs vanish. Commenters fire back: traders won’t trade by chat, others decry AI doomsaying, and several blame macro turmoil—not robots—for the recent stock slide.
A founder who spent a decade building pricey industry software (think Bloomberg-for-finance, Lexis-for-law) just declared that AI chatbots are melting the old moats—those hard-to-learn interfaces and costly onboarding that kept customers locked in. With Anthropic dropping industry plugins for its Claude “Cowork” agent, he says the chat window nukes the need for training and slashes switching costs. Cue the chaos.
The loudest clapback: traders roasting chat. One scoffed, “show me a single trader who’d ditch 36 monitors for a slow chat,” and the vibe was very “don’t take my hotkeys.” Another went full coder meme: “Imagine replacing Vim or Emacs with typing full sentences for fun”, arguing shortcuts are precise while bots waffle. Some were just exhausted: “sick and tired of the AI doomerism,” as if every spreadsheet is about to sprout a chatbot overlord.
Then came the fact-checkers. “The market slide wasn’t caused by AI,” said one, pinning the drop on politics and a huge U.S. jobs revision. Another poked holes in the stats and index counts. And, of course, the ultimate forum diss: “Was this generated by AI?”
Verdict: it’s a steel-cage match—chatbots vs. muscle memory—with Wall Street nerves, coder pride, and meme-fueled skepticism all swinging for the fences.
Key Points
- •The article reports a significant recent selloff in software and services stocks, citing large declines for FactSet, S&P Global, and Thomson Reuters.
- •Anthropic released industry-specific plugins for Claude’s Cowork, an AI agent for knowledge workers.
- •The author argues LLMs are eroding several traditional moats of vertical software while leaving others intact, reshaping value and valuation.
- •Vertical platforms (e.g., Bloomberg, LexisNexis, Epic, Procore, Veeva) historically rely on specialized interfaces, high prices, and high retention.
- •Chat-based LLM interfaces can reduce switching and onboarding costs, as illustrated by contrasts between Bloomberg Terminal workflows and LLM agent prompts, and by the author’s experiences at Doctrine and Fintool.