Breaking the Silent Barrier
Accelerating self-doubt to software competence
🔌 Watch my recent 2 hour live-stream recording with Hugo from the Vanishing Gradients that explores this topic and much more. Podcast version out shortly.
Woefully the Dunning-Kruger effect is often misinterpreted… overconfident people are likely incompetent.
A view of suspicion.
The reality of the bias is that in early learning stages people often lack the metacognitive tools to accurately assess their ability.
Much more insidious is the silent barrier.
Beginners—often actually lacking confidence—experience crippling self-doubt and fail to capitalize on the competence they already have.
The Reality in Tech
This pattern shows up all the time in Tech.
I blame it as a primary reason for why 99.5% of humans on earth can’t create software.
Despite possessing the necessary ingredients—analytical and problem solving capabilities—uncertainty and fear of failure reign.
Application may start but surrenders quickly after the inevitable hardship.
Low confidence coupled with slow feedback cycles leads to agency reduction. It prevents otherwise capable learners from:
Experimenting with new approaches
Persisting through complex problems
Contributing to discussions
Taking on challenging projects that would accelerate their growth
But…
You and I know the tide is here.
We can will digital products into existence.
Vibe coding with tools like Cursor is all the rage. It even has a wikipedia.
And like clockwork the internet does the internet.
First its in awe. And then…
It flames.
It screams.
It polarizes.
But it also misses the point.
AI is a confidence accelerator
Especially when it comes to building software.
Yes these tools can be used to showoff.
Look no further than the next three.js game on that part of twitter or an “all show no function” SaaS app created in a few hours.
These are probably cases where confidence and competence for the creator won’t soon converge.
Your actions and perceptions need to aim: 1)at accomplishing practical ends 2) at the exercise of thought 3)at maintaining a confidence founded on understanding. An unobtrusive confidence—hidden in plain sight. — X.9
But when convergence is the goal — in addition to implementation — AI can very handedly provide contextual guidance and immediate feedback.
Unlike human mentors who may be unavailable at critical moments of doubt, using AI to build software offers persistent, judgment-free assistance.
It allows the eager learner to:
Validate their thinking in real-time
Experiment safely
Break down complex problems into manageable steps
Receive personalized guidance tailored to specific challenges
This is the virtuous cycle.
One where engagement increases confidence and then competence.
It begins to mirror the “2 Sigma Effect”. Even if its not always 100% right.
When viewed this way it’s clear the primary benefit is not a replacement for human intelligence or jobs or whatever.
But rather one of the most powerful tools ever available to bridge a gap that so often rate limits our own potential.