Quiet Confidence: An Engineer's Perspective
Confidence doesn't come from speaking loudly or showing off; it stems from core competence and the ability to manage risk.
We often confuse someone who is truly skilled with someone who is simply the loudest in meetings. In the engineering world, confidence doesn’t stem from charisma; it comes from the ability to manage risk.
What is true confidence?
Many people think confidence is the ability to present fluently or dominate others in a discussion. But from a pragmatic standpoint, confidence is simply knowing the limits of yourself and the system you are working on.
It’s not about believing you’ll never make a mistake. It’s about knowing exactly what to do when a mistake actually happens. A good engineer doesn’t promise their code is perfect; they promise that if there is a bug, it will be detected and handled immediately.
The two sides of confidence at work
Loud Confidence
This is the type of person who always claims everything is easy before even looking at the system requirements. They rely on emotion, motivation, and persuasion. The problem with loud confidence is that it often lacks a factual foundation. When risks materialize, these individuals tend to blame external circumstances.
Quiet Confidence
If you are maintaining a monorepo with 200k lines of code with just four people, this is where the difference shows. Someone with quiet confidence doesn’t need to announce they are in control. They write comprehensive test cases, set up alerting systems, and always have data recovery scripts ready. They work in silence, but their results speak for themselves.
| Criteria | Loud Confidence | Quiet Confidence | Notes |
|---|---|---|---|
| Origin | Emotion, ego | Data, experience | Quiet confidence is more sustainable |
| Reaction to errors | Blame, panic | Analysis, mitigation | Most visible when the server goes down |
| Risk assessment | Often ignored | Always has a backup plan | |
| Communication | Dominant, talkative | Concise, direct |
When AI teaches us about fake confidence
We can clearly see the illusion of confidence when working with modern large language models. Versions like Claude Sonnet 4.6 or GPT-5.2 are very intelligent, but they sometimes provide incorrect answers with an air of absolute certainty.
This “hallucination” phenomenon is the technological version of loud confidence. I previously analyzed this in the post Cursor vs Claude Sonnet 4.5: Real-world code comparison. AI often doesn’t hesitate to generate logically flawed code while explaining it in an extremely persuasive manner.
(It sounds counterintuitive, but sometimes a system that knows how to say “I don’t know” provides higher value than one trying to make up an answer).
Similarly, when analyzing reading comprehension in the post Claude Sonnet 4.5 and GPT-5 summarizing Vietnamese books, the model that dares to omit details it is uncertain about is the one that delivers more reliable results. The ability to recognize one’s own blind spots is a sign of true intelligence.
How to build realistic confidence
- Separate your ego from your work: When someone points out a flaw in your design, they are criticizing the design, not you as a person.
- Build a safety net: Don’t rely on luck. Use automated testing tools and always back up your data.
- Accept temporary ignorance: Don’t be afraid to ask basic questions. It’s normal not to know a new technology; hiding your ignorance is the real problem.
Frequently Asked Questions
How can I be confident when starting out?
Don’t try to “act” confident. Focus on learning, taking careful notes, and being honest about what you don’t know. Confidence will naturally form as your competence grows.
Does being an introvert reduce confidence?
No. Introversion is just how you recharge your energy. You can absolutely be soft-spoken while making sharp technical decisions based on solid data.
Is overconfidence dangerous?
Extremely. It leads to skipping basic checks, which often causes serious, irreversible consequences in large systems.
Conclusion
To me, confidence isn’t about completely eliminating fear. The difference lies in whether you have enough data, tools, and processes to control that risk. A person who stays quiet but knows exactly what they are doing is always more reliable than someone who is loud but empty.