I’m a behavioral philosopher. I spend most of my time studying why intelligent people make decisions that, in hindsight, look obviously irrational.
In tech, this question comes up constantly.
Why do users stick with buggy software they complain about every day?
Why do engineers stop caring about the quality of code they ship?
Why do startups cling to dying products long after the data says it’s time to pivot?
Why do teams ignore security warnings until a breach makes them unavoidable?
The usual explanations—lack of incentives, bad leadership, technical debt—are not wrong. But they miss a deeper pattern.
What I see instead is something I call the digital labyrinth: a system where people remain trapped not because they lack options, but because the known cost of staying feels psychologically safer than the uncertain cost of leaving.
Known Bugs Are Safer Than Unknown Features
Product managers know this problem well.
Users ask for improvements. You build them. Adoption is low. Feedback says, “It’s great, but I went back to the old version.”
This isn’t just resistance to change. It’s a documented cognitive bias known as the ambiguity effect: people prefer options with known probabilities over ones with unknown outcomes—even when the unknown option is objectively better.
A buggy app is a known failure. Users know where it crashes. They know how to work around it. The bugs are irritating, but predictable.
A new feature, even a superior one, introduces uncertainty:
- Will it break my workflow?
- Will it cost me time to relearn?
- Will it expose me as incompetent in front of my team?
From a UX standpoint, the problem isn’t usability. It’s identity safety. People don’t abandon tools because they’re bad. They abandon tools because they threaten competence.
In tech, familiarity often beats excellence.
Why Developers Stop Caring About Code Quality
The same pattern shows up inside engineering teams.
Early-stage startups are often obsessed with craftsmanship. Engineers care deeply about architecture, security, and long-term maintainability. Then the company grows. Timelines compress. Decisions are made further from the codebase.
Slowly, something shifts.
Engineers stop pushing back.
Code reviews become superficial.
Security warnings are acknowledged, then ignored.
This is not laziness. It’s something psychologists call the agentic state—a condition where individuals stop seeing themselves as moral agents and start seeing themselves as instruments of a system.
When a developer feels like:
“I don’t own this outcome anymore”
They disengage ethically as well as emotionally.
This is why large tech organizations accumulate massive technical debt even when they employ brilliant engineers. When autonomy disappears, so does responsibility.
People don’t sabotage systems out of malice. They withdraw care when they no longer recognize themselves in the outcome.
Normalized Deviance and the Cybersecurity Trap
Cybersecurity offers one of the clearest examples of the digital labyrinth at work.
Most major breaches don’t happen because of zero-day exploits or genius hackers. They happen because of normalized deviance.
A small warning is ignored.
Nothing bad happens.
The warning becomes background noise.
Repeat this cycle enough times, and the abnormal becomes normal.
A shared password here.
An unpatched dependency there.
A “temporary” exception that lasts three years.
The system doesn’t collapse suddenly. It erodes quietly.
The danger wasn’t invisible—it was familiar.
Teams often say after a breach:
“We knew this could happen.”
But knowing is not the same as acting. Familiar risk feels manageable, even when it’s catastrophic.
Startups and the Comfort of Predictable Failure
Founders are not immune to this psychology.
In fact, they’re especially vulnerable to it.
I once worked with a business owner in a legacy market who refused to modernize his operations despite shrinking margins and declining relevance. When asked why, he said something that has stayed with me ever since:
“At least I understand how I lose money here.”
This mindset is everywhere in tech.
A startup sticks with a failing product because:
- Investors understand it
- The team’s skills are aligned with it
- The metrics, while bad, are familiar
A pivot doesn’t just change strategy—it threatens identity.
Who are we if this fails?
What are we good at then?
What happens to our story?
Known failure offers psychological shelter. Uncertain success demands reinvention.
Digital Permanence: The New Walls of the Labyrinth
In the digital age, identity is not just internal—it’s archived.
Old tweets.
Public GitHub commits.
Leaked Slack messages.
Blog posts written in a different ideological era.
This digital permanence creates invisible walls. People hesitate to evolve because their past selves are searchable.
Changing your mind now feels like contradicting a permanent record.
So people stay consistent instead of correct.
They defend outdated positions instead of updating them.
They remain loyal to systems that no longer serve them.
The labyrinth isn’t enforced by others. It’s enforced by memory.
Why Advice Doesn’t Work (and What Does)
Most tech advice fails because it targets behavior without addressing structure.
We tell teams:
- “Communicate better”
- “Take ownership”
- “Prioritize security”
- “Be user-centric”
But these are surface-level instructions.
The deeper work is structural:
- Where does autonomy actually live?
- Who bears the cost of failure?
- What identities are being protected?
- Which risks have become familiar enough to ignore?
People don’t resist change because they don’t understand it. They resist change because they understand too well what they stand to lose.
Escaping the Labyrinth Requires Meaning, Not Motivation
Motivation is overrated.
Most people are already motivated—to preserve coherence, competence, and dignity.
Real change happens when:
- uncertainty is made navigable
- identity can survive transition
- the cost of staying becomes visible, not abstract
In tech, the question isn’t:
“How do we innovate faster?”
It’s:
“What familiar failures are we protecting because they feel safer than becoming beginners again?”
Until teams confront that question, they’ll keep optimizing the maze instead of leaving it.