David Creelman

Until recently, most HR professionals had never heard of “cost of compute.” Inference cost, token cost—these sounded like IT problems. HR had enough on its plate. That is changing. As AI use spreads and organizations experiment with agents, compute cost is becoming a management issue. It may not hit every HR department this year. But it’s worth understanding now—before it shows up as a budget surprise.

Imagine a colleague says “We should use assessment A”, and you reply, “Perhaps assessment B would be better.” Ok, all is in order. Now imagine the colleague saying “We should use best practice.” What are you supposed to say now? This dilemma is a hint that something is a bit fishy about best practices.

One of the roles HR can play in AI adoption is helping leaders understand human behaviour. One human behaviour you can count on is that if a tool helps people get their work done, then banning that tool is unlikely to be effective. You might tell people not to use a certain Large Language Model (LLM), but if they need it for their work then they are likely to run it on their phone if it’s blocked by the corporate systems.