So far the best AI tool use policy I've read, specially this bit:
Your PR body should be providing context to other developers about why a change was made, and if your name is on it, we want your words and explanations, not an LLM's. If you can't explain what the LLM did, we are not interested in the change.
What about the IP issues? Not even talking about the "ethics" of "ip theft via AI" or anything, you just know a company like Microsoft or Apple will eventually try suing an open source project over AI code that's "too similar" to their proprietary code. Doesn't matter if they're doing the same to a much greater degree, all that matters is they have the resources to sue open source projects and not the other way around. If a tech company can get rid of the competition by abusing the legal system, you just know they will, especially if they can also play the "they're knowingly letting their users use pirated media that we own with their software” card on top of it.
you just know a company like Microsoft or Apple will eventually try suing an open source project over AI code that’s “too similar” to their proprietary code.
Doubt it. The incentives don't align. They benefit from open source much more than are threatened by it. Even that "embrace, extent, extinguish" idea comes from different times and it's likely less profitable than the vendor lock-in and other modern practices that are actually in place today. Even the copyright argument is something that could easily backfire if they just throw it in a case, because of all this questionable AI training.
And specially for Microsoft, they would be shooting their own foot if they were to spread Fear, Uncertainty and Doubt in the development community over the legality of the use of AI tools like Copilot, which they themselves promote and sell.
I love this approach - clear and concise rules. They are being realistic about AI created code making it into their codebase, while setting some guidelines.
13 Comments
warm@kbin.earth · 37 pts · 230d
Fuck AI, but a respectable stance.
And then you have the ShareX developer, using a screenshot of an AI response to respond to a GitHub issue.
marcie@lemmy.ml · 9 pts · 229d
based sharex dev, every foss dev should be legally allowed to shoot one issue creator each year
Evotech@lemmy.world · 7 pts · 229d
Based. It's entirely opt in.
JetpackJackson@feddit.org · 5 pts · 230d
Wow
beyond@linkage.ds8.zone · 1 pts · 229d
roofuskit@lemmy.world · 29 pts · 230d
Seems like a common sense approach.
FirmDistribution@lemmy.world · 19 pts · 230d
yeah. It's basically "don't commit poorly written code" (AI or not)
LiveLM@lemmy.zip · 13 pts · 229d
So far the best AI tool use policy I've read, specially this bit:
HiddenLayer555@lemmy.ml · 5 pts · 230d
What about the IP issues? Not even talking about the "ethics" of "ip theft via AI" or anything, you just know a company like Microsoft or Apple will eventually try suing an open source project over AI code that's "too similar" to their proprietary code. Doesn't matter if they're doing the same to a much greater degree, all that matters is they have the resources to sue open source projects and not the other way around. If a tech company can get rid of the competition by abusing the legal system, you just know they will, especially if they can also play the "they're knowingly letting their users use pirated media that we own with their software” card on top of it.
eager_eagle@lemmy.world · 8 pts · 229d
Doubt it. The incentives don't align. They benefit from open source much more than are threatened by it. Even that "embrace, extent, extinguish" idea comes from different times and it's likely less profitable than the vendor lock-in and other modern practices that are actually in place today. Even the copyright argument is something that could easily backfire if they just throw it in a case, because of all this questionable AI training.
Ferk@lemmy.ml · 7 pts · 229d
And specially for Microsoft, they would be shooting their own foot if they were to spread Fear, Uncertainty and Doubt in the development community over the legality of the use of AI tools like Copilot, which they themselves promote and sell.
kalpol@lemmy.ca · 3 pts · 230d
Well I wish the AI would fix the Schedules Direct problem
mrnobody@reddthat.com · 2 pts · 230d
If anyone needs to translate, try out
https://libretranslate.com/
babka420@szmer.info · 1 pts · 197d
I love this approach - clear and concise rules. They are being realistic about AI created code making it into their codebase, while setting some guidelines.