It will of course be theatre. Wherever they go will be LLMs too, even if there's a different social contract.
Speed cameras don't stop speeding, the death penalty doesn't stop narcotics, copyright doesn't stop LLM vendors from stealing every byte they can ingest.
But maybe, just maybe, if I hit downvote hard enough, the human programmers will have somewhere to go where LLMs aren't used.
The question is, of the three camps, which will produce the best code:
1. Positive LLM code welcomed
2. Neutral LLM code OK
3. Negative LLM code rejected.
Ignore the actual, sort every project into these categories and project. LLMs users can still operate in each but because of the potential for contribution and risk there of...
Im assuming the policy will never prevent LLM code. The policy is purely two things: self select people into these groups and determine how the people in each group will use AI.
Looks like luddites will need to switch.
It will of course be theatre. Wherever they go will be LLMs too, even if there's a different social contract.
Speed cameras don't stop speeding, the death penalty doesn't stop narcotics, copyright doesn't stop LLM vendors from stealing every byte they can ingest.
But maybe, just maybe, if I hit downvote hard enough, the human programmers will have somewhere to go where LLMs aren't used.
The question is, of the three camps, which will produce the best code:
1. Positive LLM code welcomed
2. Neutral LLM code OK
3. Negative LLM code rejected.
Ignore the actual, sort every project into these categories and project. LLMs users can still operate in each but because of the potential for contribution and risk there of...
Thats how a pragmatic decision could be made.
Re-run the hypothetical, but this time without the LLM labels. Because LLM disclosure will always be voluntary.
Im assuming the policy will never prevent LLM code. The policy is purely two things: self select people into these groups and determine how the people in each group will use AI.