Your dashboard is lying to you. Not with numbers—with confidence.
We can now generate hypotheses in seconds, pull in live data, simulate scenarios overnight, and ask an LLM almost anything. So why do some of the best teams still fail? I think it's because we've confused computation with judgment. Computing tells you what is. Judgment decides what matters.
Good judgment in 2026 looks like asking what question we're actually answering. We need to notice when our metrics become our goals, and have the nerve to ignore data that doesn't serve the decision at hand. It's the meta-skill: the ability to decide how to properly decide.
The more automated and powerful our tools, the more human and incisive our judgment needs to become. The real question isn't whether AI can help us decide—it should not! It should give us the data underlying a good decision. But the essential question is whether we still know how to form a good decision.