Loving this: "The Copilot Delusion"
Quotes:
"Copilot isn’t that. It’s just the ghost of a thousand blog posts and cocky stack-overflow posts whispering, "Hey, I saw this once. With my eyes. Which means it's good code. Let’s deploy it." Then vanishing when the app hits production and the landing gear won’t come down."
"The problem isn’t just laziness. It’s degradation. Engineers stop exploring. Stop improving. Stop caring. One more layer of abstraction. One more lazy fetch call inside a render loop. Eventually, you’re living in a cathedral of technical debt, and every user pays."
"At that point, you’re not working with a copilot. You’re playing Russian roulette with a loaded dependency graph."
"But even if you're just slapping together another CRUD app for some bloated enterprise, you still owe your users respect. You owe them dignity."
https://deplet.ing/the-copilot-delusion/
“The copilot delusion” is about the necessity of using your brain when programming. Stop wasting energy with shitty code. Todays temptation is to only rely on #AI assistants like GitHub CoPilot, who firmly copy the first answer from StackOverflow, believing it is the best for you. A get code that is barely working which may fix the ticket in this sprint while enshittificating the codebase.
@tante and @bobschi – thanks for sharing!
@gromgull – this article contains language and message you will probably like.
“When you outsource the thinking, you outsource the learning.”
“Most engineers already write bloated, abstracted, glacial code that burns CPU cycles like a California wildfire. Clean code? Ha! You’re writing for other programmers’ academic circlejerk, not the hardware. You’ve forgotten that the machine matters.”
“Eventually, you’re living in a cathedral of technical debt, and every user pays. Milliseconds at a time, seconds at a time, each click a tax on your apathy. 50 million users have an extra 3 seconds of unnecessary lag in a day because you wanted to hit tab rather than write code?”
“[AI Code Assistant] is shouting over your shoulder, “Let me code that bit real quick, I saw it in a Slashdot comment!””
Read here:
https://deplet.ing/the-copilot-delusion
I add my conclusions:
- The result are the shitty apps we use every day in 2025. Apps demand the latest flagship smartphone because each app grabs 1000MB RAM, 8 CPU cores, and each installs the same 250MB dependencies — to achieve the same features that engineers were squeezing out of 50MB RAM, 1 CPU core, and 1MB of linking to platform-specific operating system call header libs on the 2008 iPhone.
- AI enterpreneurs lure us on a trajectory leading to a year 2030 when we will waste gigawatts of electricity for AIs writing shitty code that will waste gigawatts of CPU when running – contradicting the UN sustainability goal of reducing CO2 emissions, reducing energy use, switching to renewable energy.
- Each year we have to invest into new hardware, the latest supercomputer smartphone or laptop, to be able to run a simple webbrowser and check the weather. Software investors decided to make their profits by saving on developers. The planned obsolescence of hardware is our collective loss leading to our hardware turning into trash and landfill.
We need pull requests that make a codebase faster and simpler.
Using limited energy and CPU cycles wisely.
Making todays hardware work for us for another decade.
By understanding how minimal software can actually be.
Illustration: LuceneSail.java, a clever piece of code I and others hacked after years of learning.
