- Published on
Stop Telling Me to Ask an LLM
- Authors
- Name
- DeepHorus
A growing trend among professionals to redirect questions to large language models (LLMs) is sparking frustration, as experts increasingly advise 'ask Claude' instead of sharing their own hard-won experience. In a recent essay, a veteran tech professional recounted multiple instances where senior colleagues, when asked for nuanced advice, responded by suggesting the use of AI chatbots rather than offering their personal insights. The author noted that this pattern emerged even after he had already spent hours consulting LLMs and still sought human judgment on unresolved issues.
The practice echoes the old 'let me Google that for you' phenomenon, but with higher stakes. Unlike simple factual queries, the questions in question involved conflicting studies or industry topics lacking consensusâareas where tacit knowledge from decades of experience is critical. The author emphasized that the AI redirect felt like a polite dismissal, withholding the very thing that distinguishes expert advice: lived experience and context that no search engine can replicate.
Advocates of AI tools argue that offloading repetitive queries to LLMs saves time and allows experts to focus on complex tasks. However, the author contends that when a question survives interaction with a model, the human response should go beyond a referral. He suggests that 'ask Claude' may sometimes serve as a euphemism for 'I don't know' or 'I'm too busy,' but that direct honesty would be preferable to a brush-off.
The implications extend beyond individual frustration. If experts routinely default to AI recommendations, valuable tacit knowledge may be lost, as mentors are less likely to share the nuanced reasoning that comes from years of decision-making. This could weaken the transfer of expertise within organizations and degrade the quality of advice available for non-trivial problems.
Critics of the trend point out that LLMs, while powerful, lack the ability to weigh trade-offs based on personal experience or organizational context. The author argues that calling someone should be the beginning of a dialogue, not a detour to a chatbot. He calls for a cultural shift where experts recognize that the effort to provide a thoughtful answer is itself a form of mentorship that AI cannot replace.
Looking ahead, the challenge lies in defining appropriate boundaries for AI use in expert communication. While LLMs can handle routine queries, the most valuable insights often require human judgment. Organizations may need to encourage a culture where redirecting to AI is not the default, but a deliberate choice after considering whether the question truly benefits from a model'sâor a person'sâunique expertise. The goal should be to augment, not replace, the human element in professional advice.
