Artificial Idea | AI careers · practical prompts · no hype Thursday, February 12, 2026 · Issue #55 · Prompt Tutorial
The strategy prompt
The strategy prompt: how executives use AI to think through hard decisions
Most AI prompts are designed for tasks with a correct answer. The decisions that actually define careers do not have one. Here is how to use AI as a thinking partner for the ones that matter most.
Issue #54 identified cross-functional judgment as the single capability most correlated with retention in AI-driven restructuring across every sector studied, with a McKinsey correlation coefficient more than twice that of any technical AI skill. Cross-functional judgment is the ability to make good decisions on problems that cross organisational boundaries and require integrating multiple functional perspectives without a clear precedent or defined process to follow.
That capability cannot be developed by reading about decision-making. It is developed by making decisions, reflecting on them, and developing over time the mental models and analytical habits that produce better decisions faster in novel situations. AI, used correctly, can accelerate that development by making the thinking more rigorous, the alternatives more explicit, and the assumptions more visible than they would be in unassisted deliberation.
The five prompts in this issue are the strategy prompt framework used by the senior professionals and executives who have developed the most sophisticated AI-assisted decision-making practices. They are not designed for everyday decisions that have clear right answers available through research or analysis. They are designed for the specific class of decisions where the information is genuinely incomplete, reasonable people could reach different conclusions from the same evidence, and the quality of the thinking is the primary variable in the outcome.
Those are the decisions that define careers. This is the framework for thinking through them.
The class of decision this framework is for
Before the prompts, it is worth being precise about which decisions warrant this level of analytical investment, because applying a sophisticated framework to decisions that do not require it is as costly as applying no framework to decisions that do.
The decisions this framework is designed for share four characteristics. They involve genuine uncertainty about outcomes rather than uncertainty that additional research would resolve. They have significant and lasting consequences for your professional trajectory, your organisation, or both. They involve competing priorities or values where there is no option that satisfies all of them fully. And they require a commitment that is difficult to reverse once made, which means the quality of the decision at the time it is made matters more than the ability to correct it afterward.
Strategic career decisions, significant organisational choices, investment decisions with long time horizons, decisions about which relationships to prioritise, and decisions about which opportunities to pursue at the expense of others all tend to share these characteristics. They are the decisions most people make quickly, with insufficient deliberation, because the discomfort of sitting with uncertainty is greater than the discomfort of committing to an imperfect choice.
The strategy prompt framework is designed to make sitting with uncertainty productive rather than merely uncomfortable, and to produce a decision that is better calibrated to the actual situation than the one that would have been made without the deliberation it provides.
Prompt 1: The situation framer
The problem it solves: establishing a precise, honest description of the decision and its context before any analysis begins, because the framing of a decision determines the quality of the analysis more than any other single variable.
Most hard decisions are poorly framed when they arrive. The professional facing them has already formed a preliminary view, already eliminated some options as unthinkable, and already loaded the description of the situation with assumptions and framings that constrain the analysis before it begins. The situation framer strips those constraints away and produces a description of the decision that a rigorous outside observer would recognise as complete and honest.
You are helping me frame a complex professional
decision precisely before I begin to analyse it.
The decision I am facing: [describe it as
completely and honestly as you can, including
what triggered it, what the options appear
to be, what the timeline for deciding is,
and what you already believe about
the right answer]
The context: [describe the professional,
organisational, and personal context
relevant to this decision, including
the relationships, constraints, and
history that shape the situation]
Please:
1. Restate the decision in the most precise
terms possible: what exactly is being decided,
by when, and with what consequences
for each available option
2. Identify any assumptions embedded in
my description that are doing significant
work in shaping how the decision appears,
and that would change the decision
significantly if they turned out to be wrong
3. Identify options I have not mentioned
that are available and worth considering
before the analysis proceeds, including
options that feel uncomfortable or
that I may have dismissed without
full examination
4. Identify what type of decision this is:
is it primarily a values question,
a strategic question, a relationship question,
or some combination, and what does
that classification suggest about
how it should be approached
5. Tell me the question I most need to
answer to make this decision well,
which may be different from the
question I think I am trying to answer
Do not offer any analysis of the options
yet. The framing needs to be established
before the analysis can be useful.
The instruction not to offer analysis yet is the constraint that makes the situation framer valuable rather than just a longer version of the decision description. Premature analysis anchors on the initial framing and produces conclusions that reflect the framing's assumptions rather than the actual situation. The framing step is worth completing fully before any analysis begins, and the model's natural tendency to move toward analysis as quickly as possible needs to be explicitly prevented.
Prompt 2: The options expander
The problem it solves: ensuring that the decision is made from a complete set of options rather than from the limited set that was visible when the decision first appeared, which is almost always a subset of the options actually available.
Hard decisions tend to present themselves as binary choices because the human mind is more comfortable with binary choices than with the genuine complexity of most significant decisions. The strategic choice between staying in a current role and accepting an external offer presents itself as a binary. The actual decision space includes variations of each option, hybrid approaches, timing variations, and conditional commitments that a binary framing makes invisible.
You are helping me expand the option set
for a decision I am facing beyond the
options currently visible to me.
The decision as framed: [from Prompt 1]
The options I am currently considering:
[describe each option honestly, including
what appeals about it and what concerns you]
Please:
1. Identify at least three options I have
not mentioned that are within the realm
of genuine possibility, including options
that feel uncomfortable, options that
require difficult conversations, and
options that would only become available
if I took a specific preparatory action first
2. For each option I am currently considering,
identify the variation of that option
that is most different from how I have
described it: what would it look like
if I pursued this option more aggressively,
more cautiously, more conditionally,
or in combination with elements of
another option
3. Identify whether any of the options
I am considering are actually the same
underlying choice described in different
terms, which would mean the apparent
choice between them is not real
4. Identify the option that I have most
thoroughly dismissed without examination
and make the strongest possible case
for why it deserves more consideration
than I have given it
5. Propose the option set I should actually
be deciding between after this expansion,
which may be different from the set
I started with
The goal is not to add options for
the sake of complexity. It is to ensure
that the option chosen is chosen from
a complete set rather than a convenient one.
Point four, making the strongest possible case for the most dismissed option, is the analytical move that most consistently produces the surprise in this prompt: the option that felt obviously wrong when the decision was first framed often looks less obviously wrong when its strongest case is articulated. It does not always change the decision. It always improves the quality of the decision by ensuring the rejected option was rejected on the merits rather than on the discomfort of engaging with it.
Prompt 3: The consequence mapper
The problem it solves: mapping the consequences of each option across multiple time horizons and multiple dimensions of impact, because decisions made by focusing on immediate consequences or on a single dimension of impact consistently underperform decisions made with a fuller picture of what each option produces over time.
You are helping me map the consequences
of each option in a complex decision
across multiple time horizons and
multiple dimensions of impact.
The decision and option set: [from Prompts 1 and 2]
For each option under serious consideration,
please map:
1. The immediate consequences: what happens
in the first thirty to ninety days
if I choose this option, including
the practical, relational, and
reputational consequences that would
be visible quickly
2. The medium-term consequences: what
the situation looks like at twelve
to eighteen months if this option
is chosen and things go roughly
as expected, and what it looks like
if things go significantly better
or worse than expected
3. The long-term consequences: what
this choice makes more likely and
less likely at the five to ten year
horizon, including the options it
opens and the options it forecloses
4. The relational consequences: how
this choice affects the specific
relationships that matter most to
my professional trajectory, both
in the immediate term and over time
5. The reversibility assessment: how
difficult would it be to change
course from this option at six months,
at twelve months, and at twenty-four months,
and what would changing course cost
at each point
Then identify the dimension of consequence
I am most likely to underweight in making
this decision, based on what you know
about how I have described the situation.
The final instruction, identifying the dimension most likely to be underweighted, is the output that most consistently adds value beyond what the professional would have produced in unassisted deliberation. Every professional has systematic blind spots in how they evaluate consequences, typically related to the dimension they find most uncomfortable to examine. Naming that dimension explicitly before the decision is made gives the professional the opportunity to examine it deliberately rather than discover after the decision that it was the variable that mattered most.
Prompt 4: The values clarifier
The problem it solves: identifying the values and priorities that should govern the decision and ensuring that the option chosen is consistent with them, rather than making a decision that is analytically optimal but personally inconsistent in ways that produce regret rather than satisfaction.
Hard decisions are often hard not because the analysis is unclear but because the values that should govern the analysis are unclear or in conflict. A professional who values both financial security and meaningful work will find every decision that trades one for the other genuinely difficult regardless of the quality of the analysis, because the difficulty is in the values conflict rather than in the information deficit. The values clarifier addresses that conflict directly rather than pretending the analysis can resolve it.
You are helping me clarify the values
and priorities that should govern a
complex decision I am facing.
The decision and option set: [from Prompts 1 and 2]
The consequence map: [from Prompt 3]
Please:
1. Identify the values or priorities that
are most in tension in this decision:
the things I care about that pull
in different directions across the
available options
2. Identify which of those values I treat
as genuinely non-negotiable, where
no option that violates them would
be acceptable regardless of its
other qualities, and which I treat
as important but tradeable against
sufficient gains in other dimensions
3. Identify any values I am claiming
to hold that my description of the
situation suggests I actually weight
less heavily in practice than in principle,
and where the gap between stated and
revealed preference is most significant
4. Identify the decision I would make
if I could be certain that no one
whose opinion I value would ever
know which option I chose, and
what that hypothetical tells me
about what I actually value most
5. Propose the values hierarchy that
should govern this decision given
everything I have described, and
identify which option is most
consistent with that hierarchy
Do not tell me what I should value.
Tell me what my description of the
situation reveals about what I actually value,
and whether the option I am leaning toward
is consistent with it.
Point four, the decision you would make if no one whose opinion you value would ever know, is the values clarification question with the most reliable track record of surfacing the gap between what professionals say they value and what they actually prioritise when social pressure is removed from the equation. It is not a comfortable question. It is the question most worth answering before a significant decision is made.
Prompt 5: The decision integrator
The problem it solves: integrating everything produced by the previous four prompts into a decision recommendation that is specific, honest about its uncertainty, and grounded in the full picture rather than in the partial picture available before the analysis began.
You are helping me integrate the analysis
from a structured decision-making process
into a specific recommendation and
implementation plan.
The situation as framed: [from Prompt 1]
The expanded option set: [from Prompt 2]
The consequence map: [from Prompt 3]
The values hierarchy: [from Prompt 4]
Please:
1. State the option most consistent with
the values hierarchy and best supported
by the consequence analysis, with a
specific rationale that references both
2. Identify the most significant uncertainty
in this recommendation: the assumption
or prediction that, if wrong, would
most change which option is best
3. Identify the conditions under which
a different option would be preferable,
stated specifically enough that I
would recognise those conditions
if they materialised
4. Propose a decision timeline: when
to decide, what information if any
would be worth waiting for before deciding,
and at what point waiting for more
information becomes a decision to
avoid deciding
5. Identify the first action to take
after deciding, the one that most
commits to the chosen direction
and that makes the decision real
rather than theoretical
Then give me your honest assessment
of the quality of this decision given
the analysis: is this a decision made
from a well-examined set of options
with a clear values basis, or is it
a decision that still has significant
unexamined elements that warrant
more deliberation before committing?
Be direct about the quality of the
decision rather than reassuring me
that it is sound. A decision that
is not yet ready to be made is better
discovered in this conversation than
after the commitment is made.
The final instruction to be direct about the quality of the decision rather than reassuring is the constraint that makes the integrator the most honest prompt in the framework. The natural endpoint of any deliberation process is a feeling that the deliberation is complete and the decision is ready. That feeling is not always accurate, and the model's default tendency toward encouragement needs to be explicitly counteracted by the instruction to assess quality honestly rather than to provide reassurance.
When to use the full framework and when not to
The five-prompt strategy framework takes between sixty and ninety minutes to run fully for a significant decision. That investment is justified for decisions with the four characteristics described at the opening of this issue: genuine uncertainty, significant lasting consequences, competing values, and difficult reversibility.
For decisions that lack one or more of those characteristics, a subset of the framework is more appropriate. Decisions with clear values alignment but uncertain consequences benefit most from Prompts 2 and 3. Decisions with clear consequences but unclear values benefit most from Prompt 4. Decisions that are well-framed but need a recommendation benefit from Prompt 5 alone.
The framework is not a substitute for judgment. It is the infrastructure that makes judgment more rigorous, more complete, and more consistent with the values and priorities that the professional would, on reflection, endorse rather than those that happen to be most salient in the moment of deciding.
The decisions that define careers are made in specific moments with the thinking available in those moments. The strategy prompt framework is how to make the thinking in those moments the best thinking available rather than the fastest thinking available.
The best thinking and the fastest thinking produce different decisions. Over a career, the accumulated difference between those decisions is the accumulated difference between the career that resulted and the career that could have.
Monday we are examining something that the past several issues have been building toward and that the data from the first six weeks of 2026 is now making visible: the specific professionals who are pulling ahead in the current moment and what exactly they are doing that is producing the separation. It is not the profile that the dominant narrative about AI and careers would predict, and understanding it changes what the next quarter should look like for every professional reading this.
The separation is happening now. Monday explains what is driving it.
— Team Artificial Idea
