AI is helping leaders make decisions faster, and in some cases, more accurately. But speed is not the same as wisdom, and accuracy is not the same as judgment. Effective leaders know when to trust the machine, when to trust the room, and when to trust their instincts.
Each source of guidance has strengths, and each has blind spots. The challenge is knowing which one deserves the final say in a given situation.
Trust the machine for pattern and scale
AI is strongest when the problem is data-heavy, repetitive, or too large for a human to scan quickly. It can spot patterns, summarize information, compare options, and surface anomalies far faster than we can. In that sense, AI is a powerful amplifier of attention.
Use the machine when you need:
- Fast analysis across large amounts of data.
- Early warning signs or hidden patterns.
- Multiple options or scenario comparisons.
Despite appearances, the machine does not understand meaning the way people do. It may identify correlations without understanding context and produce confident output based on weak reasoning. Leaders need to guard against mistaking correlation for causation, a theme I explored in The Human Edge: Leadership in the AI Era. Trust the machine for pattern detection, not final judgment.
A useful question is:
Is this a problem of volume, repetition, or comparison?
If yes, the machine probably deserves a seat at the table.
Trust the room for reality and resonance
The room is where leadership becomes social. A room full of smart people can catch things no algorithm will spot: political nuance, emotional undercurrents, implementation friction, and the pragmatic reality of what people will support. Group dialogue is especially valuable when the issue affects alignment, trust, or commitment.
Use the room when you need:
- Broad perspectives on a complex decision.
- Early reaction to an idea, message, or plan.
- Reality checks on feasibility and adoption.
- Buy-in from the people who will live with the consequences.
The room is not always right, of course. Groups can drift into consensus too quickly, avoid hard truths, or simply amplify the loudest voice in the room. In the best case, however, collective intelligence will surface what individual brilliance misses. Trust the room when the issue depends on shared ownership, cultural fit, or real-world adoption. For more on the power of the group, see Facts don’t move people—peers do.
A useful question is:
What will this decision feel like to the people affected by it?
If that matters, the room matters, too.
Trust your instincts for ambiguity and meaning
Instincts are often dismissed as vague or unscientific, but our instincts are actually compressed experience. They are the brain’s way of drawing on years of exposure, tacit knowledge, and subtle signals that may be hard to articulate.
Use your instincts when:
- The situation is ambiguous or novel.
- Judgment about character, timing, or fit is required.
- The “right” answer is not fully knowable.
- Something feels off, even if you cannot determine why.
That said, instincts are not infallible. They can be distorted by fatigue, fear, ego, and bias. Strong leaders do not treat instinct as the sole source of truth; rather, they consider instinct a signal worth investigating. In ambiguous situations, further investigation is warranted because we may be tempted to rationalize what we want to believe instead of what we actually know, as I explored in Dissonance, decision making, and relationships.
A useful question is:
Is this intuition grounded in real-world experience, or just an emotional reaction?
It may be worth pausing to distinguish signal from stress, especially under pressure when emotions can masquerade as insight.
A simple rule for using the three sources of guidance
- Trust the machine when the problem is large, technical, or pattern based.
- Trust the room when the problem is human, political, or requires commitment.
- Trust your instincts when the problem is uncertain, time-sensitive, or resistant to analysis.
In most cases, the best problem-solving sequence is machine first, room second, instinct last. Let AI broaden your view, let the room test your assumptions, and let your instincts make the final call when the evidence is incomplete.
The order matters because it helps prevent both overreliance on technology and overreliance on opinion.
Leadership in the AI era is not about choosing between human intelligence and machine intelligence. It is about orchestrating them wisely. Discerning leaders know when to let the machine do what it does best, when to listen to the room, and when to trust the quiet signal that says this is the way forward.
Take time to reflect
Reflection turns practice into better judgment; as you review decisions and patterns, you build the habit of learning from your own leadership choices, a discipline I wrote about in Why reflection is important.
As you face your next important decision, pause and ask yourself:
Where am I over-trusting the machine?
Where am I relying too much on the room?
When have my instincts been grounded in experience rather than anxiety?
What might change if I slowed down long enough to separate signal from noise?
References
Clear, J. (2018). Why facts don’t change our minds. James Clear.
Jennings, R. E., et al. (2022). The power of reflection for would-be leaders: Investigating individual work reflection and its impact on leadership. Personnel Psychology.
Klein, G. A. (1998). Sources of power: How people make decisions. MIT Press.
Milotich, M. (2014). Dissonance, decision making, and relationships. Claxus.
Milotich, M. (2023). Why reflection is important. LeadingWell with Mark Milotich.
Milotich, M. (2025a). The illusion of correlation: Avoiding false patterns in decision making. LeadingWell with Mark Milotich.
Milotich, M. (2025b). Facts don’t move people—peers do. LeadingWell with Mark Milotich.
Milotich, M. (2026). The Human Edge: Leadership in the AI Era. LeadingWell with Mark Milotich.
LeadingWell with Mark Milotich is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.




