June 5, 2026
Delete first, ask questions later
"Maybe later" was a feature
The hottest new productivity hack? Not building the thing everyone swore they needed
TLDR: The article argues that the best feature is often the one you never build, especially now that artificial intelligence makes it easy to create extra stuff no one really needs. Commenters split between cheering this as common sense, defending passion projects, and joking that “Maybe later” sounded like an annoying software update button.
A deceptively simple essay about not making more stuff turned into a full-on comment section identity crisis. The author’s big idea is that some of the most valuable work in software is the work that never happens: the feature that stays on the wish list, the shiny rewrite that never launches, the grand plan that quietly dies before it can become tomorrow’s headache. In plain English: sometimes skipping a project is the smartest move, especially now that artificial intelligence tools make it ridiculously easy to crank out extra code and clutter.
That take lit up readers in very different ways. One camp basically yelled, “Yes, exactly!” They said these tools make it dangerously easy to build things on a whim, only to realize later the “great idea” was just extra mess. Another camp pushed back hard, arguing this only makes sense if you see coding as a business chore. One commenter wanted more weird, free, passion projects, not fewer, and saw faster building as a win, not a warning. Then came the galaxy-brain snark: “There is still code you aren’t writing” — a tiny comment with massive eye-roll energy.
And because no internet discussion can stay on one track, someone hilariously admitted the title made them think this was about the dreaded “Maybe later” software update button that basically means “I will harass you again tomorrow.” Meanwhile, others proposed a spicy compromise: if building is now cheap, maybe the real skill is getting better at unbuilding. The mood was clear on the thread: half cautionary tale, half digital decluttering manifesto, with a side of meme-worthy annoyance.
Key Points
- •The article argues that code not written can be more valuable than code added to a repository.
- •Many backlog items remain undone because they were not prioritized highly enough by the business or engineering team.
- •The article says that after several years, many unbuilt features would likely have become irrelevant or legacy burden.
- •It states that LLMs can already generate strong code in some narrow situations, but their output is often hard to read and maintain.
- •The article warns that easier AI-assisted coding may lead teams to build low-priority features, increasing codebase complexity and API dependency risks.