The Friction Tax Was a Filter
For 25 years, developer S. Frangulov kept a drawer of untested ideas. Each one carried hope: maybe this one is the one. Testing an idea cost evenings and energy he didn't have after work. The expense acted as a natural filter.
Then he installed Claude Code. The cost of testing dropped to zero. An idea now costs an evening to test, not a month.
"So I started pulling ideas out of the drawer, one by one. And they started dying. Quietly, fast, like flies," he writes on DEV Community. "The one that was going to be huge turns out mediocre by the third evening. The next one is worse."
A month in, the realization hit: the idea was never the valuable part. The hope before testing was. Friction worked as anesthesia, keeping him from feeling the drawer is mostly junk.
The Rabbit Hole Trap
The second half of the story is funnier—and more dangerous. He sat down to post a message to Telegram. The built-in tools couldn't do it. Fine, let's write our own sender. And let's have the model run through the subscription I already pay for, not the API. And let's make the sender add its own workflows. And let's have those workflows write workflows.
By 2 a.m., he didn't have a Telegram post. He had a tiny AI newsroom full of agents, with exactly one user.
The same pattern repeats for research: "The built-in one is mediocre, so let's write our own. Properly. With blackjack and programmatic quality gates." Memory becomes its own saga: try a hundred implementations, bolt on codegraph, then something else. By morning it's a monster nobody understands. He deletes it and starts a fresh lap the next day.
Before Claude Code, writing his own sender was a few days of work, so he grabbed something off the shelf and did the job. Now "let's just write our own" is free. So he picks "our own" every time. Every single time. Forever.
The Brake Was Load-Bearing
Friction was a brake. Expensive to test an idea? Weak ones stayed in the drawer with the hope. Expensive to roll your own? You grabbed something ready and went to do the work. The barrier filtered junk on the way in and kept you out of rabbit holes on the way out.
"The whole time I called it the enemy," he writes. "Claude Code removed the friction. And the brakes with it. Now nothing slows me down. Ideas die under instant testing. Self-improvement runs for the sake of self-improvement."
By every metric he's more productive. Agents spinning everywhere, prototypes by evening, tools built to fit his hand, tests, gates. A perfect homemade sender that, for some reason, he never uses. He has never gotten so much done.
"There is just not much left to get done. The ideas are tested and buried, the dream right behind them, and all my energy pours into late-night platforms with one user."
What Seniors and Juniors Should Take Away
Frangulov asks two questions: "Seniors with a couple of decades in: was your friction load-bearing too? Or did losing it actually set you free?" And: "Juniors: you started without the friction. Does any of this land, or does it read like a guy mad at his own productivity?"
The core insight: AI coding assistants like Claude Code remove not just busywork, but also the natural constraints that kept many senior devs focused and honest. The article doesn't offer a solution—"the only thing left for me to figure out, with its help, is how to put some friction back."
For developers using AI tools, this is a warning. Productivity isn't the only metric. Without friction, you might build a lot of things you don't need and never ship what you actually set out to do.
