LLM policy

[Sandro Santilli]

On Sat, Jun 13, 2026 at 11:01:42PM +0200, Even Rouault wrote:
> > I really cannot see this as reasonable.  Everyone is all humans, and
> > extending it to LLMs seems novel and at best highly controversial.
> 
> Strong +1 here. I'm very shocked by statements that LLMs could qualify as
> "someone". Someone obviously involves a biological entity originating from
> human cells. If we start to acknowledge rights to bots, then humanity is in
> high danger. Especially since ethical AI doesn't exist (there might be lab
> experiments around that, but in practice that's not the one people use when
> they use commercial models). So you're currently dealing with things
> explicitly designed to make human users addict. There are numerous
> testimonies of events where "agents" start leaving the framework in which
> they were initially thought to be restricted and play according to their own
> "intents", and given the high dependence of our societies to computer
> networks, this is deeply concerning.

I agree machines shouldn't have rights. Not as of 2026 at least :)

I hope we humans will handle to retain our rights as we loose the power,
as that's what have been happening over time, from DVD players destroying
your right to skip advertisement, to EBook readers destroying your right to
borrow books to friends, to CI system destroying your right to accept a
contribution your peers consider good, to proprietary code based services
destroying your right to own modify and distribute the tools you use.

LLMs are not the problem, ownership of LLMs is.

--strk; 

  Libre GIS consultant/developer 🎺
  https://strk.kbt.io/services.html
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 659 bytes
Desc: not available
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20260618/e401707d/attachment.sig>