The people building AI's future just admitted they can't keep up with it either.
The Summary
- Cat Wu, head of product for Claude Code at Anthropic, told Lenny's podcast that users are experiencing FOMO from the relentless pace of AI releases, forcing people to "check Twitter every single day" to stay current
- The traditional monthly or quarterly release cycle has collapsed. AI labs now ship so fast that even their own products overlap and confuse users.
- Wu's diagnosis: the industry created a treadmill it doesn't know how to slow down, and the solution is better onboarding, not fewer releases
The Signal
Anthropic's product lead just said the quiet part out loud. The AI race isn't just leaving users behind. It's leaving the builders behind too. Wu admitted that Anthropic ships features so fast they sometimes overlap with each other. Translation: even the people inside the lab can't map their own product surface area anymore.
This matters because we're watching the collapse of product development as we knew it. Software used to move in quarters. You'd ship, measure, iterate. Users had time to form mental models. Now? Claude Code shipped updates this month that users complained degraded output quality. Before those complaints could resolve, new features landed on top. The feedback loop is breaking.
"People feel this need to check Twitter every single day to see what the absolute latest thing is."
But here's the real signal buried in Wu's comments. She frames the problem as user education. "I would love people to feel like they can just open these tools. The tools will educate or teach them what they want to know." That's not a solution. That's punting. The issue isn't that users need better onboarding. It's that the entire industry is stuck in a Red Queen's race where shipping speed is the only metric that matters.
Look at the landscape Wu describes:
- AI labs competing on coding tools
- Big Tech rolling out agent frameworks
- Startups launching overlapping products weekly
- Users unable to evaluate what actually works before the next thing drops
The conventional wisdom says this pace is necessary. If you don't ship, you fall behind. But Wu accidentally revealed the cost: products that overlap with themselves, users who can't keep up, and a growing gap between capability and usability. Anthropic is trying to win by building everything, everywhere, all at once. So is OpenAI. So is Google. The result is a fragmented mess where even power users feel lost.
The Implication
If the people building AI agents admit they're overwhelmed by the pace, what does that tell you about mainstream adoption? We're not headed toward a future where AI seamlessly integrates into work. We're headed toward choice paralysis and tool fatigue. The winners won't be the labs that ship fastest. They'll be the ones that figure out how to make their tools feel stable and trustworthy while everything else spins faster.
Watch what Anthropic does next. If they slow down, it's a signal the race is maturing. If they keep shipping and just add better tooltips, you'll know they're stuck on the treadmill too.