"They have a pretty straightforward business, make GPU cards and sell them."
They do, but that's not the (full) story here. Companies tend to easily migrate upwards, to a higher volume and/or higher profit margin market, and hardly (if ever) in the opposite direction. The painful restructuring necessary to enable this kind of reverse change is also damaging to the company brand, culture, and self-perception. If they ever get in such position, they may of course recover, but I wouldn't bet money on that.
The only reason for a source code to be is for humans to read it, bun when the source code gets churned (by AI agents) in too large of a quantity for any human to realistically read and analyze, then what's the point of having a source code in the first place? Generating binary directly simply makes sense. Working with binary does, even when a human is involved, as long as there's an AI helper as well. The human simply can ask the AI assistant to explain whatever logical aspects behind the binary code and instruct the AI agent to modify the binary code directly, if necessary. That may be scary and not easy to accept. Going further with this idea, even the written text may become "too costly to work with" when there will be an AI agent to verbally or graphically serve the human with whatever informational aspect of a given text that could be of interest in a given situation.
LLMs are trained on source code, so that's what they can (barely) write. Decompiling is a -lossy- action which means that training directly on the output would have much less information and would be a nightmare if one (human or llm) needs to debug.
>I cannot conceive of a way that any form of healthy life, does not want to expand it's resources to improve future outcomes, especially one that is maximally optimized for thinking.
"Then you have a very limited imagination."
This is not about imagination. Given the space of possibilities to act or evolve, if mentioned expansion cannot somehow be ruled out, then it makes sense for it to be assumed (with enough time, for whatever time can mean in this context) as a certainty, even for non-organic "life".
If the current RAM heavy buyers, the AI powerhouses investors, don't get the into a profitable state of business, then sustaining this rhythm "for a long time" becomes impossible. It won't matter much that "fab allocation is booked years out" if the client that expects the goods goes out of business, doesn't it? I, for one, don't find convincing hints that this free AI crazy partying will go on for long, so then what gives?
I think we both agree on that, just on different timelines. I think there's enough VC money to drag this out long enough for it to really hurt. And the longer they do, the more the entire planet becomes dependent on AI companies staying afloat lest the whole economy collapse.
> If the current RAM heavy buyers, the AI powerhouses investors, don't get the into a profitable state of business, then sustaining this rhythm "for a long time" becomes impossible.
Exactly that is the problem with the "pork cycle" we are seeing [1] - there aren't that many manufacturers and ODMs around nowadays for RAM, storage, CPUs and GPUs. The ecosystem was so much more vibrant even 10 years ago. When the AI bubble collapses, it will take the entire world's economy down the drain, and I think that quite a few of the brands we have now will be extinct after this iteration.
"Microsoft's GitHub is hosting the leaked source code (which probably got sucked into Copilot and every other AI under the sun as a result)." "However, I don't think copyright lawyers will care. «They're also committing a crime» doesn't mean you're free to do what you want. That applies especially in ReactOS vs MS, because if ReactOS succeeds, it will compete directly with Microsoft."
There's also such thing as being responsible (for an outcome), which in case of litigation means being culpable. Microsoft here is the sole actor that has any control on the GitHub Copilot, on what it was fed with, and thus - on its output (which would be the base of their accusation if they sue). How do you imagine such a case could be made to look like it would have any legal standing?
"In January 2006, concerns grew about contributors having access to leaked Windows source code and possibly using this leaked source code in their contributions. In response, Steven Edwards strengthened the project’s intellectual property policy and the project made the difficult decision to audit the existing source code and temporarily freeze contributions."
The allegations have been taken seriously and since then the procedure for accepting contributions include measures to prevent such further events from occurring. If you or anyone else happen to have any plausible suspicion, then please report it to the ReactOS team, otherwise keeping alive this kind of vague and uncertain connection between some Windows code leakage and ReactOS fits the very definition of FUD: https://en.wikipedia.org/wiki/Fear,_uncertainty,_and_doubt Please stop.
"if Microsoft attempts to sue (themselves?) over their own Copilot tool injecting their own copyrighted code into a user's codebase"
Such an attempt can't make sense, given that the model used by ReactOS is in Microsoft's control and thus Microsoft alone is the one responsible for the model's behavior. They won't sue, thus much is clear.
But, if any such model got fed with leaked code, then how is this a specific open-source project's problem and not of all projects (either open-source or private) that got to ever use that model?
Then, (having thought this just now) how can an argument relying on (legally) undisclosed information be used against anything public? Isn't the onus on the party having the undisclosed information to prove that it preceded the public one? How can that precedence be trusted by an independent judging party if the undisclosed information (source-code and systems managing that source code) is and always has been in the hands of the accusing (thus biased) party?
"it probably scrapped the multiple git repository of Windows leaked source code. In which case it would ABSOLUTELY undermine the project's ability to say it's a clean room implementation"
If an LLM model has been fed leaked code, then that is a general problem for that model and for its use for anything. Singling out its use for an open-source project and denouncing that as a potential problem while otherwise keeping quiet about it just makes no sense. Just take legal action against the model if there's anything plausible to warrant that, don't weaponize it against open-source projects.
All LLM have probably as they scrape github, and there are still to this day multiple Windows XP source code live on it (I won't give links but they are pretty easy to find). And I'd bet there is way more than just windows leaks on there...
They do, but that's not the (full) story here. Companies tend to easily migrate upwards, to a higher volume and/or higher profit margin market, and hardly (if ever) in the opposite direction. The painful restructuring necessary to enable this kind of reverse change is also damaging to the company brand, culture, and self-perception. If they ever get in such position, they may of course recover, but I wouldn't bet money on that.
reply