Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.project > #13721 > unrolled thread

Re: Do we need a conflict of interest policy?

Back to article view | Back to linux.debian.project

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: Do we need a conflict of interest policy? "G. Branden Robinson" <g.branden.robinson@gmail.com> - 2025-02-07 17:50 +0100
    Re: Do we need a conflict of interest policy? Pierre-Elliott Bécue <peb@debian.org> - 2025-02-08 01:10 +0100
      Re: Do we need a conflict of interest policy? Charles Plessy <plessy@debian.org> - 2025-02-08 02:10 +0100
        Re: Do we need a conflict of interest policy? Timo Röhling <roehling@debian.org> - 2025-02-08 10:00 +0100
    Re: Do we need a conflict of interest policy? "Jonathan Dowland" <jmtd@debian.org> - 2025-03-06 12:20 +0100

#13721 — Re: Do we need a conflict of interest policy?

[Multipart message — attachments visible in raw view] — view raw

[M-F-T set to -project; see https://cr.yp.to/proto/replyto.html ]

At 2025-02-07T17:08:44+0100, Pierre-Elliott Bécue wrote:
> I'm not sure I'd expect more from people than them saying who they
> work for or defend.
> 
> I'd not, eg, expect some Canonical employee to refrain voting a GR
> because they might be biased. We all are biased.

I think this is a non sequitur.  Yes, we are all human (except for the
LLM that wrote Charles's draft policy), and therefore we all have
cognitive biases.

That does _not_ imply that we all have conflicts of interest.

Biases are often unconscious, whereas conflicts of interest are always
known to the person who has them.

Regards,
Branden

[toc] | [next] | [standalone]

#13726

[Multipart message — attachments visible in raw view] — view raw

"G. Branden Robinson" <g.branden.robinson@gmail.com> wrote on 07/02/2025 at 17:41:41+0100:

> [M-F-T set to -project; see https://cr.yp.to/proto/replyto.html ]
>
> At 2025-02-07T17:08:44+0100, Pierre-Elliott Bécue wrote:
>> I'm not sure I'd expect more from people than them saying who they
>> work for or defend.
>> 
>> I'd not, eg, expect some Canonical employee to refrain voting a GR
>> because they might be biased. We all are biased.
>
> I think this is a non sequitur.  Yes, we are all human (except for the
> LLM that wrote Charles's draft policy), and therefore we all have
> cognitive biases.
>
> That does _not_ imply that we all have conflicts of interest.
>
> Biases are often unconscious, whereas conflicts of interest are always
> known to the person who has them.

I don't think it's always clear to them.

But I get your point and I agree that there's a difference.

I'm not sure that, considering that we are around a 1000 people here, it
really is an issue.

-- 
PEB

[toc] | [prev] | [next] | [standalone]

#13727

> "G. Branden Robinson" <g.branden.robinson@gmail.com> wrote on 07/02/2025 at 17:41:41+0100:
> > Biases are often unconscious, whereas conflicts of interest are always
> > known to the person who has them.
 
Le Sat, Feb 08, 2025 at 12:55:38AM +0100, Pierre-Elliott Bécue a écrit :
> I don't think it's always clear to them.

At my workplace, we have "rules for conflicts of interest management",
and in of them aims at "preventing situations in which [the researchers]
may be perceived to have a conflict of interest".

So here we have three different scopes:

 - conflict of interest is concious to individuals,
 - conflict of interest can be unconcious to individuals,
 - conflict of interest can be outside the control of individuals.

As we see in this discussion, there is a considerable disagreement
within Debian about the very definition of what a conflict of interest
is.  I think that this makes us ill-equipped to to handle situations
where one developer expresses the opinion that another developer has a
conflict of interest.

We are not going to solve that overnight, but I feel that we need to
agree on the minimum, which is:

 - it is not a taboo,
 - it is not an attack,
 - it is not a misconduct.

This is all that I have been aiming for when I started this discussion.

By the way, for the full disclosure, and also to counter the argument
that using ChatGPT to write a text in a language that I am not
comfortable with was an act of carelessness in which I dumped
low-quality text produced with no effort, here is the inputs I gave:

Prompt 1:

Please draft a very brief conflict of interest statement for a
non-profit organisation such as Debian, that has a very diverse
membership, and diverse understanding, acceptation and experience of the
concept of conflict of interest. The main point to address is that
people have employers or sponsors which themelves can be donators, have
interest, or be impacted by Debian's decisions. The policy should advise
on how to handle the information and when to refrain from participating
to a decision, and also how to raise the possibility of a conflict of
interest without offending the person, etc.

Not satisfied with the results, I added prompt 2:

Can it explain a bit more what a conflict of interest is, acknowledge
that opinions may differ on that point, and also say that there is
nothing bad at all of being in a situation of conflict of interest?

I was then satisfied that what I read was what I wanted to write, but
better written that if I would do it myself.

Some parts are not perfect, but if there is appetite we have a lot of
talented people here who can make it better.

We do not have to conclude soon.  We can give the chance to everybody,
including the vast majority of people who read but do not write on this
list, to think about it, take the time it needs to be exposed to
situations in their life where the question of "what is a conflict of
interest" has some relevance, before we try to progress further.

Have a nice week-end,

Charles

-- 
Charles Plessy                         Nagahama, Yomitan, Okinawa, Japan
Debian Med packaging team         http://www.debian.org/devel/debian-med
Tooting from home                  https://framapiaf.org/@charles_plessy
- You  do not have  my permission  to use  this email  to train  an AI -

[toc] | [prev] | [next] | [standalone]

#13729

[Multipart message — attachments visible in raw view] — view raw

Hi Charles,

* Charles Plessy <plessy@debian.org> [2025-02-08 10:02]:
>We are not going to solve that overnight, but I feel that we need 
>to agree on the minimum, which is:
>
> - it is not a taboo,
> - it is not an attack,
> - it is not a misconduct.
>
+1

>By the way, for the full disclosure, and also to counter the 
>argument that using ChatGPT to write a text in a language that I am 
>not comfortable with was an act of carelessness in which I dumped 
>low-quality text produced with no effort, here is the inputs I 
>gave:
>[...]
For the record and notwithstanding my humorous take on model 
collapse, I happen to think that it is absolutely okay to use LLMs 
like this, because their primary function is text generation, and 
they do this very well. Arguably better than me, at the very least.

And let's not forget we are talking about a draft here, reviewed and 
refined by someone with domain knowledge at that. This is not 
someone replacing critical Linux kernel code with untested Copilot 
drivel, nor an attempt to cheat an exam or fake a scientific 
publication.

>Have a nice week-end,
Likewise!

Cheers
Timo

-- 
⢀⣴⠾⠻⢶⣦⠀   ╭────────────────────────────────────────────────────╮
⣾⠁⢠⠒⠀⣿⡁   │ Timo Röhling                                       │
⢿⡄⠘⠷⠚⠋⠀   │ 9B03 EBB9 8300 DF97 C2B1  23BF CC8C 6BDD 1403 F4CA │
⠈⠳⣄⠀⠀⠀⠀   ╰────────────────────────────────────────────────────╯

[toc] | [prev] | [next] | [standalone]

#13752

I stumbled across this Debian Wiki page I'd forgotten about. We have had 
a voluntary Register of Interests here since 2017:

https://wiki.debian.org/RegisterOfInterests


-- 
Please do not CC me for listmail.

👱🏻	Jonathan Dowland
✎	 jmtd@debian.org
🔗	https://jmtd.net

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.project

csiph-web