Hi André,
On 2026-05-29 at 20:42-03:00, André Batista wrote:
> sex 29 mai 2026 às 11:21:10, Nguyễn Gia Phong enviou:
> > Let's take another example, if C gives us a car P from manufacturer S,
> > while insisting that it doesn't happen ever despite S saying
> > P fails to brake 1% of the time, would you trust C and ride P around?
> > What would any court thinks if you hit someone with P
> > and blame C's assurance? Of course, it would be a different scenario
> > if C withheld the car's model.
>
> So, as I understand it, you are implying that in this situation, if
> manufacturer S advertises selfdriving car model P as the solution to
> car accidents as they remove the human error factor ("the safest car
> ever sold, you must buy it now!") and then sells a billion cars which
> would cause ten million of accidents throughout the world every 3
> months they could avoid any liability by virtue of inserting a clause
> in their contracts saying "we can't promise that car model P won't go
> berserk, run accross children on streets, run off clifs or
> selfexplode"? And that if I had bought one of these cars I would
> become liable for these accidents[?]Yes, if you were aware of that clause and decided to drive it anyway, and to be pendantic, it closer to "we are aware of car model P _occasionally_ going berserk [...]". On 2026-05-29 at 20:42-03:00, André Batista wrote: > and even after selling it to you I'd remain liable in [your] place? No, then I'd be liable instead, when I chose to drive it. S should be held liable for all its frauds, but we citing its false advertisements as an excuse to do something we know to be wrong is just pure malice. On 2026-05-29 at 20:42-03:00, André Batista wrote: > sex 29 mai 2026 às 11:21:10, Nguyễn Gia Phong enviou: > > If big tech boosters want to change to status quo and incorporate > > LLM outputs along with their risks into free software projects, > > they should present their evidence and do the convincing. > > I object to singling out and name calling users of these services. > Also, my idea was just that: leave a narrow opening through which > people can present their evidence and try to convince the rest that > this tech has some usefulness and that the risks can be mitigated. I was calling people boosting services can only be provided by mega tech corporations and maybe some governments by their actions, and I assume they actually truly believe such adoption is for the greater good. I'm happily awaiting their evidence besides their personal experience with the SaaSS. On 2026-05-29 at 20:42-03:00, André Batista wrote: > qui 28 mai 2026 às 17:54:14, Nguyễn Gia Phong enviou: > > There is practically zero inclusion of LLM output in Guix codebase. > > Attempts to introduce legal, technical, and moral risks, etc. > > to the project should preach for acceptance, not the other way around. > > sex 29 mai 2026 às 11:21:10, Nguyễn Gia Phong enviou: > > I am disappointed that the discussion here treats LLMs as inevitable > > and caters to their users instead of securing our software freedom. > > If one day all courts worldwide rule that all LLM outputs > > are under the public domain, then it's not too late for another GCD > > reverting the their ban anyway. > > I'm not so sure about this. Even though Ludo claims this GCD is only > about the "Detailed Design", the motivation section shows a great > deal of concerns over this tech that goes way beyond copyright. So > say the courts decide that LLM outputs are public domain, so what? > Will we then ignore all other concerns and jump the fascist, climate > denialist, human craft enemies bandwagon as if none was the matter? > Wouldn't that be a bit hypocritical? We cater to those concerns to > support us when convenient, but drop them once they become obstacles? To be fair, I did mention _technical, and moral risks_, when they no longer exists, the it is possible to discuss lifting the ban. I'm taking a very standard conservative stand here: don't threaten the legal, technical, social and economical safety of the project without sufficient evidence backing the proposed progress. I'm happy to receive news from Denis that GNU is asking to > not include new code/data generated by LLMs On 2026-05-29 at 20:42-03:00, André Batista wrote: > Also, what does it mean to secure our software freedom? To me it > means having user rights recognized by law, constitutions and human > rights treatises. It means abolishing intellectual property rights > over source code. It means making it unlawful to furnish binary > blobs without source code and documentation. It means making it > unlawful to provide hardware without publishing its datasheets and > designs. It means having public oversight over any SaaSS offered > to the public. I agree with all these, but we can only work with what we have, which currently is copyleft. Giving up on the possibility of enforcing it without the mentioned utopia being present is just unwise IMHO. On 2026-05-29 at 20:42-03:00, André Batista wrote: > In order to get there, we'll need massive popular support and this > support will not be gained by secluding ourselves on a free software > island of novena laptops. It will not be gained by calling people > "useds". It will not be gained by shutting people out unless they > mirror us in all ways of behaving and thinking. Thanks, this made me chuckled a bit, and I feel called out. While we do need as much support as possible, I'm wary of policies that appear popular yet alienating our movement's exist base. There's no going back after welcoming LLM into our project with open arms, and honestly it's become incredibly daunting to dodge big tech's influences across most free software projects lately. See also _stricter is less popular_ by Alexandre Oliva: <https://blog.lx.oliva.nom.br/2026-02-22-stricter-is-less-popular.en.html> Best wishes, Phong
signature.asc
Description: PGP signature
