Self-sabotage does not always look self-destructive. For capable people, it can look responsible: one more analysis, one more system, one more tool, one more round of polishing before the work ships.
That is why Charlie Munger’s inversion framework has become useful to me. Instead of asking only, “What should I do to succeed?” I have started asking a more uncomfortable question: If I had everything I needed and still underperformed, how would I have done it?
Inversion is a decision-making technique that works backward from failure. You define an outcome you do not want, identify the behaviors, assumptions, and risks most likely to produce it, then avoid planting those seeds in the first place.
That sounds simple. Applied honestly, it can expose a different problem: sometimes the obstacle is not a lack of intelligence, knowledge, technology, or opportunity. Sometimes it is the way we use them.
What did Charlie Munger mean by inversion?
Warren Buffett gave one of the clearest public descriptions of Munger’s approach in Berkshire Hathaway’s 2009 shareholder letter. The section is titled “What We Don’t Do.”
“All I want to know is where I’m going to die, so I’ll never go there.”
Buffett wrote that Munger’s thinking was inspired by mathematician Carl Jacobi’s practice of inverting difficult problems. Berkshire then described how that logic affected real decisions: avoiding businesses they could not evaluate reliably and arranging their finances so they would not become dependent on outside liquidity during a crisis.
The point was not pessimism. It was failure prevention before optimization.
Munger’s broader work on judgment followed the same pattern. In his 1995 speech “The Psychology of Human Misjudgment,” he described seeing recurring patterns of irrational behavior and building a practical system for recognizing them. The lesson I take from that speech is not simply “be smarter.” It is to study the predictable errors that can defeat otherwise intelligent people.
Why inversion hit me differently now
I have spent years getting better at adding things.
Knowledge. Technology. Marketing methods. AI tools. Workflows. Buyer simulators. Operating systems. Research. Frameworks. New opportunities.
That ability has created real leverage. I naturally look for connections between the customer, revenue, people, process, technology, and evidence. When something works, I want to understand why and make it repeatable.
But a strength can create its own failure mode.
I can procrastinate with high-quality work.
Not by doing nothing. By doing sophisticated work that is useful, defensible, and sometimes genuinely good, but still is not the work that determines the outcome.
I had already started seeing this pattern in my own work. When I tested HubSpot’s Loop Marketing framework, the first decision I made was not to build another marketing operating system around it. I kept the parts that improved work already in motion and left the rest out.
I reached a similar conclusion while designing my AI workflow. Access to more capable tools eventually stopped being the constraint. The harder problem became deciding what each system should own and where human judgment still mattered.
And in a recent article about decision making and artificial prerequisites, I argued that a process can become so responsible-looking that it protects delay instead of the outcome.
Munger’s inversion framework ties those observations together. It changes the question from “What else can I add?” to “What am I already doing that makes the result less likely?”
AI makes sophisticated self-sabotage cheaper
This matters more to me in the AI era because AI has made it much faster for me to create another layer of work.
With AI, I can produce a long strategy, a new operating framework, a scorecard, a research plan, several alternative workflows, a risk matrix, and another round of copy revisions much faster than I could before.
That is extraordinary leverage when the work matters.
It is also an extraordinary way to become busy without becoming consequential.
AI has turned my systems instinct into both leverage and risk. I can create useful structure faster than ever. I can also create unnecessary structure faster than ever.
That means the scarce capability is moving. The bottleneck is increasingly less about whether I can generate another option and more about whether I can judge which option deserves to exist.
I inverted my own failure pattern
I asked myself a specific question:
If I reached the end of the next year and had dramatically underperformed my potential, despite having the experience, tools, knowledge, and opportunities I needed, what would most likely have caused it?
The answers were not dramatic. That is what made them useful.
1. I kept refining work that was already good enough to ship.
Quality matters. So does the point where another revision stops changing the business outcome. My failure mode is not careless work. It is letting refinement continue after the material risk has already been addressed.
2. I built systems before repetition proved a system was needed.
I like making good work reusable. The inversion is asking whether the repeated problem exists yet. If I am not careful, I can build a whole system for a process I have only done a few times.
3. I kept too many legitimate futures alive at the same time.
This one is harder because the opportunities are not bad. They are often credible. The problem is that five good futures can consume enough attention to weaken the one that needs to become real now.
4. I let low-consequence work feel like progress.
Research can feel productive. Planning can feel productive. Improving a system can feel productive. The inversion test asks whether the work protects income, creates cash, advances a near-term decision, removes a material risk, or produces evidence that changes what happens next.
5. I protected old commitments because I had already invested in them.
Projects, assets, identities, and systems can outlive their usefulness. The time already spent does not make the next hour automatically valuable.
These are not abstract weaknesses I pulled from a productivity book. They are patterns I can recognize in my own work.
That is what makes inversion useful. It gives me a way to convert self-awareness into operating constraints.
One revenue goal made the distinction harder to fake
I recently changed the way I define progress on a revenue goal.
Pipeline does not count. A proposal does not count. Expected commissions do not count. An invoice that has not been paid does not count.
I call those things possibility money. They can matter operationally, but they are not the outcome.
Collected cash counts.
That definition is inversion applied to measurement. Instead of asking how I can make the dashboard look encouraging, I ask how I could fool myself into believing I had achieved the result before the result existed.
Once that failure mode is visible, the metric becomes much harder to game.
There is research behind the broader “consider the opposite” idea
Munger’s inversion framework is a mental model, not a scientific law. But decision research supports a narrower mechanism behind it.
In a 1984 study published in the Journal of Personality and Social Psychology, researchers Charles Lord, Mark Lepper, and Elizabeth Preston found that instructing people to consider possibilities opposite to their current beliefs reduced biased judgment more effectively than simply telling them to be objective.
Gary Klein’s premortem applies a related idea to projects. A team imagines that the project has already failed and works backward to generate plausible reasons. The point is to surface risks that optimism, hierarchy, or commitment might otherwise hide.
Neither method proves that inversion will produce the right answer every time. They support something more modest and useful: deliberately generating opposing or failure-based explanations can expose information that ordinary forward planning misses.
The Second-Pass Audit is inversion applied to AI work
The same logic is already built into something I use called the Second-Pass Audit.
Generative AI is very good at producing something that looks finished. A strong answer can create its own psychological pressure to accept it.
The Second-Pass Audit reverses that direction. Instead of asking, “Does this look good?” it returns to the source and asks what the answer omitted, distorted, overclaimed, or underweighted.
The underlying principle is the same: generation creates a candidate; verification earns acceptance.
That is a more useful relationship with AI than either blind trust or reflexive skepticism.
How I use inversion without turning it into another system
Start with the actual outcome. If the outcome is vague, inversion only creates vague fears. Define what success means in observable terms.
Write the failure sentence. Assume the result did not happen. Ask what you most likely did, ignored, repeated, or tolerated that contributed to the failure.
Separate ruin from ordinary mistakes. Not every error deserves a new control. Focus first on behaviors capable of creating unacceptable or repeated downside.
Build the smallest useful guardrail. A bedtime alarm, a cash definition, a source comparison, a spending limit, a decision rule, or a hard stop can be more useful than another operating framework.
Then go do the positive work. Inversion does not choose the goal for you. It protects the goal from predictable self-inflicted damage.
The goal is not to become smarter everywhere. It is to stop repeatedly doing the few things that make your intelligence irrelevant.
What inversion is changing for me
For years, learning more created enormous leverage for me. It still does.
But the question has changed.
I am less interested in collecting another smart idea simply because it is smart. I want to know whether it changes a decision, improves work already in motion, prevents a meaningful failure, or deserves to become repeatable.
That fits the progression I have been using for my own development: Learn → Apply → Transform.
Inversion belongs underneath it, not beside it as another operating system. It is the failure-prevention question that keeps learning from becoming accumulation, application from becoming activity, and transformation from becoming another project.
So the question I am carrying forward is straightforward:
How would I predictably screw this up?
Find the few answers capable of doing real damage. Stop feeding them. Then let good decisions and time compound.
Frequently asked questions about inversion and self sabotage
What is inversion in decision making?
Inversion is a decision-making method that approaches a problem backward. Instead of only asking how to create the desired result, you identify the behaviors, assumptions, and risks most likely to cause failure and work to avoid them.
What did Charlie Munger mean by “invert”?
Munger used inversion to focus attention on avoidable failure. Berkshire’s 2009 shareholder letter shows the idea operationally through decisions about business selection, liquidity, and risks the company was unwilling to tolerate.
How can inversion reduce self sabotage?
Inversion makes recurring failure patterns explicit before they repeat. For a capable professional, that might mean over-refining finished work, adding unnecessary systems, keeping too many priorities active, or measuring activity instead of the actual outcome.
Is inversion the same as a premortem?
No. Inversion is a broader reasoning principle. A premortem is a specific planning technique that asks a team to imagine a project has already failed and generate plausible reasons for that failure.
When should leaders use inversion?
Use inversion when the downside matters, when repeated mistakes are possible, when optimism may hide risk, or when a team has many attractive options and needs to identify what could quietly defeat the desired outcome.
Sources and further reading
- Berkshire Hathaway 2009 shareholder letter, including “What We Don’t Do.”
- Charlie Munger, “The Psychology of Human Misjudgment” transcript.
- Lord, Lepper, and Preston, “Considering the opposite: a corrective strategy for social judgment”, Journal of Personality and Social Psychology, 1984.
- Gary Klein, “Performing a Project Premortem”, Harvard Business Review, 2007.
