Deliberate Intentional Practice with AI
AI's perceived ineffectiveness stems from insufficient deliberate practice, not fundamental limitations—mastery requires the same intentional experimentation as learning a musical instrument.
Summary
When engineers say AI doesn't work for them, the crucial question is: from which identity are they speaking? Are they sharing experiences from a particular company's legacy codebase, or have they tried AI at home with deliberate, intentional practice?
Companies with ancient codebases and proprietary patterns lack the training data AI needs. That experience is entirely understandable. But engineers whose only AI exposure is struggling with legacy code may never discover what the tool can actually do.
The Instrument Metaphor
The TB-303 bass synthesizer failed commercially at launch. Years later, someone started twisting knobs in strange and wonderful ways—and new genres of music emerged.
AI works the same way. People who get the most from it don't just pick up a guitar, experience failure, and conclude the instrument is broken. They play. They approach with the intention of not achieving much, just strumming with an open mind to discover something new.
The COBOL Experiment
One evening on Zoom, drinking margaritas, Huntley and a friend wondered: can AI program COBOL? Moments later, they built a calculator. Then a Reverse Polish Notation calculator. Then unit tests in COBOL.
Their brains racing, they pushed further: what about an RPN calculator using emojis as operators? Does COBOL even support emojis?
One way to find out: ghuntley/cobol-emoji-rpn-calculator
That exact moment—riffing on possibilities without any goal beyond exploration—is deliberate intentional practice.
The Employer/Employee Dynamic
Context windows are small. Some enterprise codebases will never fit. But that's a company problem, not an employee problem.
There was a time when engineers left companies that wouldn't adopt AWS. Employees exchange skills for money, and the industry advances. Engineers who don't upskill risk becoming unable to exchange their skills at market rate.
AI not working at a particular company doesn't mean AI doesn't work. Hope isn't lost—context windows will grow, semantic analysis will improve, build system integration will mature. But employees waiting for their employer to figure it out may fall behind those who practice at home.
Key Quote
"In the circles around me, the people who are getting the most out of AI have put in deliberate, intentional practice."
Connections
- deep-work - Newport's framework for deliberate practice applies directly here: mastering AI requires the same focused, distraction-free experimentation that builds any complex skill