Claire Vo is fed up. In a world where new AI models drop weekly, her exhaustion isn't from keeping up with raw intelligence. It's from the sheer volume of verbal fluff, specifically what she calls "Claude slop" – the verbose, apologetic, and heavily hedged prose churned out by Anthropic models like Opus 5.

Key Takeaways

  • The "Slop" Tax: Claire Vo describes Anthropic models' verbose, hedging in-chat prose as "Claude slop," causing her "blood to boil" due to its exasperating and often nonsensical nature for busy founders.
  • GPT's Direct Counter: Vo praises OpenAI's GPT models for their directness and "bullet points" approach, highlighting a significant difference in user experience and readability.
  • Performance vs. Presentation: Despite her personal frustration with Opus 5's chat style, Vo acknowledges the model ultimately excelled in her "How I AI" benchmark for front-end design and prototyping tasks.
  • Hidden Costs: The core issue isn't raw AI capability, but the invisible friction and time sink of sifting through excessive verbiage just to get to the useful output.

The Hidden Tax of AI Politeness

For ambitious builders and founders, every minute counts. Sifting through overly cautious, verbose AI responses isn't just annoying – it’s a tax on productivity. Claire Vo's frustration with Opus 5 hit a boiling point over its chat style. She details her near-meltdown:

“I cannot read Claude's slop anymore. I am losing my mind with Claude's slop. And the Claude's slop is Claude's slopping, baby.”

Vo isn't just complaining about bad grammar. She's highlighting a fundamental user experience problem. Imagine asking for a direct solution and getting a preamble full of caveats, disclaimers, and self-analysis. Opus 5, she found, possessed a surprisingly neurotic and human-reliant personality, spending valuable chat real estate telling her what she wanted to hear rather than simply delivering the goods.

GPT's Brutal Efficiency

Vo draws a sharp contrast with OpenAI's GPT models, which she credits with fixing this problem early on. “OpenAI fixed it by just being like we are bullet points and we are product manager talk. We're very direct,” she explains. This isn't just about aesthetics; it's about efficiency.

While Claude would offer a “meta-analysis of like, 'Plus I'm telling you what you want to hear,'” GPT would cut to the chase with direct warnings: “be careful about me being confident, me being wrong, privacy, outdated information, bias, emotional authority, and over-dependence.” One is a conversational dance, the other a crisp memo. For a founder racing to build, the latter is often gold.

The Productivity Paradox of Over-Politeness

Here's the twist: despite Vo's exasperation with the "Claude slop," Opus 5 ultimately excelled in her specific "How I AI" benchmark for front-end design and prototyping tasks. This creates a love-hate dilemma. The model's underlying intelligence was potent, but its output presentation created significant friction.

This paradox reveals a critical insight: raw intelligence isn't the only metric for an AI's utility. The interface of that intelligence – how it communicates and delivers information – directly impacts a founder's ability to extract value. A powerful engine with a terrible UI still slows you down. The conversational hedging, meant perhaps to build trust or manage expectations, instead becomes a productivity blocker, turning a quick query into a slog.

What to Do With This

Review your AI workflow this week. For high-volume, output-critical tasks like coding, content generation, or data analysis, explicitly prompt your AI with brevity requirements such as "Respond in bullet points only," or “Give me the direct answer, no preamble.” If a model consistently delivers "slop," consider A/B testing a more direct alternative like GPT for those specific jobs, even if another model shows stronger raw intelligence in other areas. Your time is too valuable to waste on AI's conversational filler.