We use cookies to improve your experience on our site and to show you relevant advertising. By clicking "Accept All", you consent to the use of ALL cookies. You can also manage your preferences or learn more about our privacy practices.
AI can make the work look better without making the decision clearer.
by Drew Robbins
4 min read
Choose What Matters
AI can widen the field and speed up the work, but human judgment still has to name the real decision before a team wastes energy on the wrong problem.
AI can make your work look sharper while your decisions get weaker.
You covered every angle, and somehow no one knows what to do next. AI helped you prepare and made sure nothing got missed, but nobody ever got clear on the decision in front of you. Now the room has more information and less direction.
That is what feels so strange about AI at work. The tools can generate so much so quickly that a team can leave with better language, more options, and less grip on what matters most. Your value is in catching that drift before the work starts pulling everyone deeper into the wrong problem.
That is why discernment matters so much right now. The World Economic Forum's 2025 skills outlook says analytical thinking remains the top core skill for employers. When answers come fast, the person who stands out is the one who can sharpen the question, narrow the field, and name what deserves commitment.
The same thing shows up in the research on AI performance. Dell'Acqua and colleagues found that consultants using AI did much better on tasks inside the tool's capability frontier, but performance dropped when they used it outside that frontier and leaned too hard on the model's recommendation. As Dell'Acqua put it, people would "kind of switch off their brains and follow what AI recommends," which was more likely to be wrong. AI can help you move faster and define options. It still cannot decide for you what deserves resources, attention, or commitment.
That makes sense when you remember what these tools are built to do. They are generation models. They tend to agree, add, and keep going, even when what you need is restraint or a harder question. A roadmap meeting can end with twice as many plausible initiatives and half as much clarity about which one actually deserves resources. You have seen it when a model takes your premise, tells you it is right, and gives you more. Even negative instructions can still feed it language that pulls the output in the wrong direction. The model will keep generating. You still have to choose.
Before you ask for more output, name the decision that actually matters. Put that on the table before your team makes another deck, another doc, or another list. Do that first if you want to protect your attention, your team's energy, and the part of the work that still needs a human being to see clearly.
You cannot make progress without making decisions.
— Jim Rohn
What keeps getting your best attention right now, and is it actually the decision you most need to make?
Try This
Before your next prompt, meeting, or draft review, write one sentence that names the real decision you need to make.
Notice What Happens
Watch how much easier it is to judge options once the problem is clear enough to decide.
Share or Reflect
Name one task this week where you were generating more when you really needed to narrow.
Keep Going
Build a habit of asking what deserves commitment before you ask for speed.
If this resonates, share with your network to help more people protect their judgment in a workplace full of options.
References
Dell'Acqua, F., McFowland III, E., Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K. C., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2026). Navigating the jagged technological frontier: Field experimental evidence of the effects of artificial intelligence on knowledge worker productivity and quality. Organization Science. https://doi.org/10.1287/orsc.2025.21838
