Open Source After Cheap Contributions

Open source runs on an asymmetry: contributing took effort, so a submitted patch carried a signal that someone cared enough to do the work. Maintainer attention was the scarce resource, and the effort required to contribute rationed demand for it.

Cheap generation removes the rationing. The scarce resource is unchanged.

The immediate pressure

Volume rises, quality distribution widens. More contributions, including many that are plausible, well-formatted, and subtly wrong or unnecessary. The plausibility is what costs maintainers time — an obviously bad patch is closed in seconds.

Issue reports get harder to triage. Detailed, well-written reports for problems that aren't problems, or that the reporter didn't actually investigate.

The effort signal is gone. A polished PR used to mean someone spent an evening on it. It no longer distinguishes.

⚠️ Maintainer burnout was already the sector's chronic problem. This pushes directly on it, and the people affected are volunteers.

What maintainers are doing about it

Patterns that appear to work, most of which raise the cost of contributing in ways that select for engagement rather than for effort:

Require an issue and a discussion before a PR. Moves the expensive filtering earlier and forces a human exchange before code exists.

Require a failing test with any bug fix. Demonstrates the problem is real and understood. Cheap for someone who genuinely found a bug, and a real filter otherwise.

Ask contributors to explain the change in their own words, including what they considered and rejected. Someone who understands their contribution can do this; someone who submitted output can't go deep.

Explicit AI-contribution policies. Ranging from disclosure requirements to restrictions on certain categories. ✅ Disclosure with a requirement to demonstrate understanding seems more workable than prohibition, which is unenforceable.

Automate what's mechanical. Tests, linting, coverage, format — every check a machine performs is maintainer attention preserved for judgment.

💡 The reframe worth making

The value a contributor provides was never the code. It was a person who understands a problem, cares about it, and will be around to answer questions about their change.

That was always the real contribution; the code was evidence of it. So the filters that work are the ones that test for understanding and engagement rather than for effort:

• Can you explain what you did and why?
• Did you consider the alternatives?
• Will you respond when this needs adjusting in three months?

Those questions were implicitly answered by the effort of contributing. Now they need asking directly.

What gets better

Not only pressure:

Genuine barriers drop for real contributors. Someone who understands a problem but struggles with an unfamiliar codebase's conventions can now contribute. That's a real accessibility gain and it produces contributions that wouldn't have existed.

Tedious maintenance work gets cheaper. Dependency updates, mechanical migrations, test coverage, documentation fixes — the work maintainers least want to do.

Non-native speakers can contribute more equally, since polished English is no longer a barrier.

→ The gain is real and it lands on the same channel as the problem, which is why blanket prohibition costs something.

✅ For contributors

If you want your contribution to be welcome:

• Open an issue and discuss first. Always, for anything non-trivial.
• Understand what you're submitting. Be able to explain every line and defend the approach.
• Include a failing test for any bug fix.
• Say how you produced it. Honesty is better received than most people expect.
• Stay engaged. The contribution isn't the PR; it's being there for the follow-up.
• Don't submit speculative fixes for problems you haven't confirmed. This is the single largest source of wasted maintainer time.

The takeaway

Open source rationed maintainer attention through contribution effort, and that rationing is gone while the attention is unchanged. The filters that work test for understanding and engagement — discuss first, explain in your own words, include a failing test, stay for the follow-up — because those are what the effort was proxying for. And the accessibility gain is real, which is why the answer is better filters rather than closed doors.