A software engineer's recent experiences reveal a fundamental distinction in how we seek answers: some questions are purely transactional while others are inherently social. When encountering an unfamiliar word, he realized he wanted to discuss it with a chef friend rather than simply look up the definition. This social dimension of knowledge-seeking contrasts sharply with programming questions that demand immediate factual answers.

These two question types—relationship-based versus transactional—often get conflated in discussions about AI replacing human work. The author illustrates this with a Python programming problem about handling None values in pandas DataFrames, which he solved efficiently using an AI chatbot. For such technical queries with clear right answers, AI tools excel. But relationship-based questions serve different purposes: they facilitate social connection, allow sharing of perspectives, and provide the context that makes facts meaningful.

The distinction becomes critical when examining claims that AI will replace strategy consulting or other white-collar professions. Strategy work isn't about finding correct answers but rather about trust, judgment, and human relationships. Clients seek advisors whose opinions they respect, not just factual accuracy. The cathartic value of explaining situations to another person and feeling understood cannot be replicated by AI systems, regardless of their sophistication.