When 'Just Ask Claude' Becomes a Polite Way to Dodge a Hard Question

The author makes a distinction that gets lost in the reflexive push toward AI tools: the value of asking an experienced human isn’t information retrieval, it’s judgment. When she calls a senior colleague to ask which of five conflicting studies he’d personally trust — the thing three decades of watching decisions go wrong taught him — being told to ‘ask Claude’ misses the point entirely. She had already spent hours with the model before ever reaching out to a person. The question she’s left with is the one that survived that process.

This isn’t the old LMGTFY dynamic of lazy people offloading trivial searches onto strangers. It’s more like asking a friend with shared taste for a restaurant they’d actually pick, not a top-ten list. The redirect to an LLM, she argues, has quietly become the polite version of ‘I don’t know,’ ‘I’m busy,’ or ‘I’d have to think about it’ — and any of those honest answers would be more useful than a pointer back to the tool she already exhausted.

The deeper point is about the cost of being the person others call. Real expertise — lived, hard to articulate, impossible to search — takes attention and effort to share, and not everyone can spare it on a busy day. That’s fair. But ‘ask Claude’ doesn’t save a step when the question has already cleared the model; it just withholds the one thing the machine can’t offer.