[conspire] Discussion: Using LLMs the Right Way: 10/1/2025 7pm Eastern Daylight time

[Rick Moen]

Quoting Ivan Sergio Borgonovo (mail at webthatworks.it):
> On 10/2/25 10:44 AM, Rick Moen wrote:
> 
>> I don't think that makes any sense.  The failure Deirdre described had
>> >very little to do with the particular SCM's semantics or data model.
>
> really?

I'm certainly not an expert, but have been using SCMs nearly as long as
they've existed and published a popular listing and comparison of all
SCMs available for Linux: http://linuxmafia.com/faq/Apps/vcs.html
I've kicked around Hg (Mercurial) just a little, and cannot see that the
problem described was an emergent effect of Hg-specific properties.

What happened was _of course_ not inherent or inevitable in an LLM 
having such a role.  However, the fact that the LLM had the power to
carry out uncommanded, serious damage, and nobody at the firm took
measures, at least not effective ones, to prevent such carnage, is
troubling.

It suggests, among other things, that LLMs in that scenario can and do
carry out non-deterministic behaviour that is highly destructive.  Of
course, we already knew that, from many other examples.  Personally, as a
Unix user, I have a strong preference for tools that function in highly
predictable ways, such that, even if I shoot myself in the foot, I can
do forensics and say "Ah, and _there_ is why I should have known that
would happen."

[big snip]

> So you're saying that I'm responsible for resource waste because I
> asked a LLM to show me a template on how to drive a PWM?
> I bet I'm also responsible for microplastics because I didn't spend
> enough time to deconstruct a fucking package made of 5 different
> materials glued together.

You know, I saw your response during the night, and decided to delay any
comment because I didn't want to sound irritable.  So, I put it off, and
re-read what you said a few times, and yet the awkward fact remains:
Your comment was completely non-responsive to what I said.

Nothing I wrote was in any way critical of, or addressed explicitly or
implicitly, ro the social responsibilites of one Ivan Sergio Borgonovo,
honoured mailing list participant in Italy.   The actual _purport_ of my 
comment is reasonably clear from its wording, I think.

Having said that, I suspect my only graceful way forward, in this
subthread, is to just ask to please not continue, and express regret 
that this did not work out better.