Decision Making for Product Leaders with a Drop of AI

This was my lighting talk at Product Jam on June 12, 2026. I realized with 2 minutes before my presentation that it needs to last no longer than 8 minutes + 2 for Q&A. I had provisioned 15 minutes for the talk. Imagine my surprise.

This deck is inspired by two amazing courses that I had at during Stanford LEAD: Decision Making and Neuroscience of Exemplary Leadership.

Decision Making Slides.pdf

Download PDF • 1.47MB

The Main Concepts - Decision Making for Product Leaders

CONCEPT 1: THE SPEED-STRESS TRAP

The acceleration of work (faster building, planning, deciding) combined with

AI-driven volume creates chronic FOMO. Chronic FOMO produces sustained cortisol

elevation. Sustained cortisol causes dendritic atrophy (~16% reduction in

prefrontal cortex spine density after 3 weeks, per Harvard Medical School research).

The prefrontal cortex governs logical reasoning, complex analysis, and emotional

regulation. When it is impaired, the amygdala dominates — producing fear-based,

reactive decisions.

CONCEPT 2: COGNITIVE SURRENDER

Source: Wharton study, 2026 — "Thinking Fast, Slow, and Artificially"

Definition: The moment when an AI model does not merely execute a task but makes

the decision, and the human adopts that decision as their own without recognizing

that the transfer of agency has occurred.

Risk: The person believes they are thinking. They are actually inheriting the

LLM's output as their own judgment.

Author's position: AI should augment decision-making, not replace it. Offloading

decisions to AI atrophies the critical thinking muscle over time.

CONCEPT 3: DECISION TYPES — SIMPLE vs COMPLEX

Simple decisions: reversible, low impact, low cognitive cost.

Rule: act fast, use heuristics or instinct, do not over-analyze.

Reason: over-analyzing simple decisions consumes cognitive capacity needed for

complex ones.

Complex decisions: high stakes, often irreversible, multi-variable.

Rule: do not offload to AI. Use AI as an augmentation layer within a structured

analytical process. Instinct is valid only after analysis — as a signal of

confidence, not a shortcut to it.

CONCEPT 4: THE ANALYTICAL TOOLKIT

Derived from: 3 Stanford courses on decision-making (analytical + psychological),

ongoing reading and testing.

Core principles:

- Start by identifying objectives clearly before evaluating options.

- Avoid binary framing. Move from 2 alternatives to 3–5.

- Use pre-mortem analysis (especially in groups) to surface non-obvious risks.

- Build decisiveness as a skill: the ability to commit to a decision and live

with it matters more than finding the theoretically optimal choice.

CONCEPT 5: MENTAL STATE AS A DECISION VARIABLE

Framework source: Professor Baba Shiv (Stanford GSB) — Type 1 / Type 2 thinking.

Note: distinct from Kahneman's System 1 / System 2.

Type 1 thinking: defensive, fear-driven, triggered by poor sleep, stress,

cortisol overload.

Type 2 thinking: creative, confident, enables bold and well-reasoned decisions.

Condition for Type 2: good sleep, physical activity, low chronic stress.

Author's insight: AI cannot substitute for the neurochemical state required to

make good decisions. Lifestyle conditions the quality of judgment.

CONCEPT 6: CORE THESIS

In a world where everyone uses the same AI models and executes faster than ever,

speed is no longer a differentiator.

Judgment is the differentiator.

Judgment is trainable.

The more decision-making is delegated to AI, the less the human judgment muscle is trained — and the less the individual trusts their own reasoning over time.

Decisions are the primary domain humans control. Not outcomes — but the process by which decisions are made.