I think that's largely fair comment. Having a battle over an issue like this isn't really my thing - hence the talk about everyone making up their own mind. If there'd been one other report on what was happening, I'd have just shared that rather than write about it. But nobody else was talking about it, even though a lot of people in the community have pretty strong feelings about AI. So, I wrote this piece to clarify to myself what I felt (OK, it doesn't feel fun to play MegaMek any more) and published it because it seemed like other people would want to know.
There's no "other report" because this is a complete non-story. Your pearl clutching on this topic misrepresents the project maintainers work and tries to drum up controversy out of nothing.
Aside from things slug out in the courts, there has been a long tradition of code-copying when it comes to development, and AI development tools are simply a faster way of assembling snippets into a workable project. They jump from code to prose or art is much more severe and one that may be where the legal line gets drawn
That is, in part, because extracting AI-influenced code from professional products is going to be infeasible at this point, not least due to having to answer the question of what percentage constitutes violation? All code undergoes value-add review (and edits, zero percent or more) before deployment. So how is that hair split?
Ultimately it will end up with individuals making value judgements regarding their comfort, however if they are not themselves developers on the team delivering free Battletech they perhaps ought to consider that devs like max progress in min time with min effort, too. It is, after all, the devs who are paying for everything.
I'd love to know why you seem to have forgotten to mention the way ML-assisted code is tested and integrated into the wider project. You also fail to mention the portraits that have been in the game for years, which are obviously AI-generated, and the way these came about. You ignore the review and testing process, instead projecting your fear of ML tools onto a non-profit, volunteer project, whilst raising false equivalences as "concern". Hell the main developer that uses ML wrote this wiki page, which is literally the 4th page on the wiki and immediately visible when it's opened (Very shady stuff for an open source volunteer project, I know). https://github.com/MegaMek/megamek/wiki/AI%E2%80%90Assisted-Code-in-MegaMek. Kinda hard to fearmonger about it when the person using it seems to hold these principles, though, so maybe it was inconvenient to your narrative? Would love to know your reasons. Hopefully, this article doesn't discourage people from using a fantastic piece of software that has had over 20 years of love poured into it, over such a knee-jerk reaction.
You’re welcome to your opinion but just know I found this to be one of the least persuasive things I’ve read in years and I believe anyone who reads this will be less sympathetic to our common Ai-resisting position.
Stuff like this destroys the credibility of legitimate resistance to actual crimes like theft of intellectual property and even likenesses. Instead, you make us all look like both Ivory Tower and Luddite.
This is an open source project with incredible amounts of rote work that has zero benefit to creativity or human contribution, as it is merely recoding the same info in a new format, which can’t be changed due to existing code that interacts with it. It is not something copyrightable, nor is it trained to do this off of copyrighted material. If Techbros wanted to strawman us, it would look like this.
You have crashed right into the other sides best argument, probably provoked reflexive backlash from the BT community for criticizing a common beloved project severely, and as a result, made future pushback to AI more difficult, with people next time very likely to remember how unreasonable our camp was this time.
Are you trying to win a debate and convince people or are you looking to enforce some kind of social expectation with pressure to conform? Seems a lot more like the latter
You are entitled to post whatever you wish. Just know the only side of this debate being damaged by this is our mutual one.
I'm sorry this is just dumb. I think there are battles to fight on AI use, but code for an open source fan project is the very worst one to pick.
I think that's largely fair comment. Having a battle over an issue like this isn't really my thing - hence the talk about everyone making up their own mind. If there'd been one other report on what was happening, I'd have just shared that rather than write about it. But nobody else was talking about it, even though a lot of people in the community have pretty strong feelings about AI. So, I wrote this piece to clarify to myself what I felt (OK, it doesn't feel fun to play MegaMek any more) and published it because it seemed like other people would want to know.
Normal service will resume next week.
There's no "other report" because this is a complete non-story. Your pearl clutching on this topic misrepresents the project maintainers work and tries to drum up controversy out of nothing.
Please see our official response - https://megamek.org/2026/01/18/Addressing-AI-Tools-in-MegaMek-Development.html
This browbeating posturing is absolutely pathetic.
Aside from things slug out in the courts, there has been a long tradition of code-copying when it comes to development, and AI development tools are simply a faster way of assembling snippets into a workable project. They jump from code to prose or art is much more severe and one that may be where the legal line gets drawn
That is, in part, because extracting AI-influenced code from professional products is going to be infeasible at this point, not least due to having to answer the question of what percentage constitutes violation? All code undergoes value-add review (and edits, zero percent or more) before deployment. So how is that hair split?
Ultimately it will end up with individuals making value judgements regarding their comfort, however if they are not themselves developers on the team delivering free Battletech they perhaps ought to consider that devs like max progress in min time with min effort, too. It is, after all, the devs who are paying for everything.
I'd love to know why you seem to have forgotten to mention the way ML-assisted code is tested and integrated into the wider project. You also fail to mention the portraits that have been in the game for years, which are obviously AI-generated, and the way these came about. You ignore the review and testing process, instead projecting your fear of ML tools onto a non-profit, volunteer project, whilst raising false equivalences as "concern". Hell the main developer that uses ML wrote this wiki page, which is literally the 4th page on the wiki and immediately visible when it's opened (Very shady stuff for an open source volunteer project, I know). https://github.com/MegaMek/megamek/wiki/AI%E2%80%90Assisted-Code-in-MegaMek. Kinda hard to fearmonger about it when the person using it seems to hold these principles, though, so maybe it was inconvenient to your narrative? Would love to know your reasons. Hopefully, this article doesn't discourage people from using a fantastic piece of software that has had over 20 years of love poured into it, over such a knee-jerk reaction.
imagine spending the time it took to write this "takedown" about (checks notes) an open source project and contributing to the project instead?
A useless hit piece that helps nobody.
You’re welcome to your opinion but just know I found this to be one of the least persuasive things I’ve read in years and I believe anyone who reads this will be less sympathetic to our common Ai-resisting position.
Had some more thoughts:
Stuff like this destroys the credibility of legitimate resistance to actual crimes like theft of intellectual property and even likenesses. Instead, you make us all look like both Ivory Tower and Luddite.
This is an open source project with incredible amounts of rote work that has zero benefit to creativity or human contribution, as it is merely recoding the same info in a new format, which can’t be changed due to existing code that interacts with it. It is not something copyrightable, nor is it trained to do this off of copyrighted material. If Techbros wanted to strawman us, it would look like this.
You have crashed right into the other sides best argument, probably provoked reflexive backlash from the BT community for criticizing a common beloved project severely, and as a result, made future pushback to AI more difficult, with people next time very likely to remember how unreasonable our camp was this time.
Are you trying to win a debate and convince people or are you looking to enforce some kind of social expectation with pressure to conform? Seems a lot more like the latter
You are entitled to post whatever you wish. Just know the only side of this debate being damaged by this is our mutual one.