[Wikimedia-l] Re: AffCom Statement on Wikimedia Slovakia vs KuboF

2021-10-28 Thread Matej Grochal
Dear Mehman

thank you for this message.
Best wishes to everyone.

Matej Grochal

On Thu, Oct 28, 2021 at 9:34 PM Mehman Ibragimov 
wrote:

> Hello Everyone,
>
> AffCom is disappointed by the recent allegations surrounding the
> Wikimedians of Slovakia (WMSVK) User Group, and the personal attacks
> against Manavpreet Kaur, a full-time contributor to the movement both
> professionally and personally.
>
> While it is not our practice to comment on matters before the Conflict
> Intervention sub-committee, out of respect for the privacy of the parties,
> in the interests of transparency, we believe it is necessary to respond to
> the recent emails posted to this list. AffCom has been engaging with Michal
> Matúšov and the Wikimedians of Slovakia User Group since May 2020. The
> dispute in question arose from a section of the 2018 Wikimedians of
> Slovakia annual report.
>
> In order to understand what happened and to determine the best course of
> action, AffCom met the User Group representatives and Michal Matúšov
> separately. After the meetings, AffCom reviewed the information and
> documentation to identify the issues and offer ways to remedy the
> questioned section in the 2018 report.
>
> To reach a consensus with the parties, AffCom explored ways that the User
> Group could voluntarily remove the disputed material. The best option was
> for Michal Matúšov to provide receipts of the disputed section for the User
> Group to review, and consequently use as supporting evidence to facilitate
> the removal of the disputed section. AffCom took this route because it
> lacks the mandate and the ability to demand the removal of content from
> User Group reports.
>
> Despite several requests, over several months, Michal Matúšov has been
> unable to provide the receipts. Instead, he requested AffCom to derecognize
> the Wikimedians of Slovakia User Group and to notify him in advance so that
> he could step in and save the group. Towards this demand, AffCom clarified,
> and still continues to clarify, that there are no specified bylaws or
> governance guidelines for User Group affiliates. This is because User
> Groups are the simplest affiliate model with no requirements for regional
> non-profit status, Board or bylaws. This structure allows groups to
> function as recognized bodies. Therefore, evaluating performance on
> unspecified standards would be unfair.
>
> It is important to note that AffCom can review implementation of bylaws if
> they have been approved by the committee. In this case, as the bylaws are
> not a mandate, they were never reviewed or approved, and as a result,
> AffCom cannot undertake office action on the grounds of bylaw violation.
>
> While we empathize with all the parties involved, we are highly concerned
> at the way this matter has been addressed in a public forum. This is
> especially so because we have clarified, to all parties, the formal route
> and redress options available. We are also extremely concerned by the
> extent of bullying that has occurred, across a number of forums. While we
> may have different views, we should strive to respect and treat one another
> with dignity and decency.
>
> As we continue to explore ways on how best we can resolve this impasse,
> AffCom wishes to request all parties to assume good faith in these
> discussions (in public and private). Our next step will be to retry
> mediation to determine the best outcome.
>
> Finally, Michal Matúšov has been the primary applicant for the User Group
> Esperanto to become a Thematic Organisation. As the conflict with WMSVK
> escalated through 2021, we made the decision mid-year to pause that
> application to try to keep Esperanto out of the WMSVK conflict. Affiliates
> involved in conflict mediation need to trust in confidentiality through the
> process. We notified Michal Matúšov, not the wider Esperanto community,
> because he was the primary applicant with whom we had been corresponding.
> We recognized that this was not fair on the Esperanto community and have
> since suggested that, until the conflict between WMSV and Michal Matúšov
> is resolved, another member of Esperanto User Group takes over the
> application to move it forward. Any further discussions on this matter will
> only be attended on official AffCom channels.
>
> In the meantime, we would like to offer our sincerest apologies to
> Manavpreet Kaur, Matej Grochal and all of those affected or upset by these
> accusations and discussions.
>
>
> *Kindly,---*
>
>
>
> *Mehman Ibragimov*
> *Vice-Chair*
>
>
> *Affiliations Committee
> <https://meta.wikimedia.org/wiki/Affiliations_Committee>*
> ___
> Wikimedia-l mailing l

[Wikimedia-l] Subject: Re: Manavpreet Kaur's role in AffCom's issues

2021-10-26 Thread Matej Grochal
Dear all

this is Matej Grochal, chair of Wikimedians of Slovakia and one of the
parties in the conflict. I am just coming into the discussion as I was not
aware of it from the beginning and was not a member of this mailing list.

Let me offer you my input on this matter.

About a year ago, Mr Matúšov and I (representing WMSK) agreed on mediation.
AffCom was to be the mediator. We had an individual call in December 2020
with AffCom but WMSK and I pushed for having a mediation call of both
parties and AffCom as the mediator. After some coordination during spring
2021, AffCom called off the mediation just as it seemed that it was going
to go ahead. The reason given was that both sides do not agree to the terms
of mediation. We emailed back for clarification while restating our
commitment to proceed with it. Around the time I reached out to Mr Matúšov
who seemed perplexed that the mediation had been cancelled and committed
himself as well. Not having a clear understanding of the situation, I
contacted AffCom who replied that mediation would not proceed due to
differences already mentioned. Since our commitment was clear, we wanted to
know what happened. They informed us that Mr Matúšov was of a different
mind about mediation and so it could not go ahead. (Not very surprising
since there are two parties to the mediation). This was somewhat perplexing
still so I kept in touch with Mr Matúšov and made sure we are on the same
page about what mediation means. We were. However, I also came to
understand that he was indeed not ready for mediation. The same conclusion
was reached from further communication with him and AffCom during the fall
of 2021.
Even if AffCom did not state explicitly that it was Mr Matúšov who
cancelled the mediation, they would have every right to do so. His mind set
did not allow it to proceed and he effectively cancelled the process. The
matter was further complicated by the fact that despite waiting for
mediation, Mr Matúšov kept posting his grievances in various places around
the community after having agreed not to do so with AffCom. I will let you
judge whether this is acting in good faith.
To sum up: while AffCom could have been somewhat faster and more clear on
certain issues, I would not blame AffCom for providing false information. I
would also appreciate if accusations of providing false behaviour were less
lightly thrown.

Thank you

Matej Grochal
___
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/Q2UMIWCKS3KYJAEYPF6DIXZEV2POIQ4X/
To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org

[Wikimedia-l] Re: [Wikimedia Announcements] Ukraine's Cultural Diplomacy Month: We are back in 2022!

2022-02-23 Thread Matej Grochal
Dear all

I am with Lodewijk on this one.
Be well and may peace come soon.

Matej

On Thu, Feb 24, 2022 at 7:24 AM effe iets anders 
wrote:

> The list archive probably has an answer to your question on 9 feb 2021.
>
> But regardless, writing about Ukrainian culture is never more prudent than
> when it is under threat. Whatever one's personal opinion on this war, I
> don't think anyone will deny that Ukrainian cultural heritage is at risk?
> Asking people to go outside and take photos, that may be a bit much right
> now, but this is a writing exercise. That is the beautiful thing about
> Wikimedia: we document things regardless of how much we 'like' them.
>
> Lodewijk
>
> On Wed, Feb 23, 2022 at 9:59 PM 4nn1l2 <4nn1l2.w...@gmail.com> wrote:
>
>> This is a legitimate concern as several users have dared to ask questions
>> as you can see at
>> https://meta.wikimedia.org/wiki/Talk:Ukraine%27s_Cultural_Diplomacy_Month
>>
>> We need more transparency regarding this decision. Granted, there was a
>> campaign last year around this time, but was it advertised at a global
>> scale? At least, I can't remember that.
>>
>> On Thu, Feb 24, 2022 at 8:28 AM Ariel Glenn WMF 
>> wrote:
>>
>>> This campaign was conducted last year at around the same time:
>>> https://meta.wikimedia.org/wiki/Ukraine%27s_Cultural_Diplomacy_Month_2021
>>> Let's keep any unwarranted speculations off of this list please.
>>>
>>> Ariel Glenn
>>> ar...@wikimedia.org
>>>
>>> On Thu, Feb 24, 2022 at 5:55 AM 4nn1l2 <4nn1l2.w...@gmail.com> wrote:
>>>
 The timing of this campaign is of real concern and not prudent, I
 think, especially for those of us who strive for neutrality.

 On Thu, Feb 24, 2022 at 2:47 AM Samuel Klein  wrote:

> Thanks for sharing -- a nice idea and gorgeous page; nice to see the 
> *wikigap
> challenge* model proliferating.
>
> On Wed, Feb 23, 2022 at 6:00 PM Valentin Nefedov <
> nefedov.valen...@gmail.com> wrote:
>
>> Hello, dear Wikipedians!Wikimedia Ukraine, in cooperation with the
>> Ministry of Foreign Affairs of Ukraine and Ukrainian Institute, has
>> launched the second edition of writing challenge Ukraine's Cultural
>> Diplomacy Month , which lasts from 17
>> February to 17 March 2022. The campaign is dedicated to famous Ukrainian
>> artists of cinema, music, literature, architecture, design and cultural
>> phenomena of Ukraine that made a contribution to world culture. The most
>> active contesters will receive prizes.
>>
>> We invite you to take part and help us improve the coverage of
>> Ukrainian culture on Wikipedia in any language!
>>
>> Sincerely,
>> Valentyn Nefedov a.k.a. Renvoy
>> ___
>> Please note: all replies sent to this mailing list will be
>> immediately directed to Wikimedia-l, the public mailing list of the
>> Wikimedia community. For more information about Wikimedia-l:
>> https://lists.wikimedia.org/mailman/listinfo/wikimedia-l
>> ___
>> WikimediaAnnounce-l mailing list --
>> wikimediaannounc...@lists.wikimedia.org
>> To unsubscribe send an email to
>> wikimediaannounce-l-le...@lists.wikimedia.org
>>
>
>
> --
> Samuel Klein  @metasj   w:user:sj  +1 617 529
> 4266
> ___
> Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org,
> guidelines at:
> https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
> https://meta.wikimedia.org/wiki/Wikimedia-l
> Public archives at
> https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/IBIAITNZ4M42ND7L7743PTHS5X22GEHM/
> To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org

 ___
 Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org,
 guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
 and https://meta.wikimedia.org/wiki/Wikimedia-l
 Public archives at
 https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/I3TQ4NVWD4VI3FQTKX53C7YNEFNJGYS2/
 To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org
>>>
>>> ___
>>> Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org, guidelines
>>> at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
>>> https://meta.wikimedia.org/wiki/Wikimedia-l
>>> Public archives at
>>> https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/GCKL7THS7CWFKWEAWDPYHMRX3BBTXTTW/
>>> To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org
>>
>> ___
>> Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org, guidelines
>> at: 

[Wikimedia-l] Re: Join the new Movement Strategy Forum community review

2022-06-13 Thread Matej Grochal
Dear all

I quite agree with Galder here. We should focus on making our own spaces
more inclusive and easier to use rather than jumping to various external
providers for this and that. Let's not forget that existing volunteers and
staff also have to learn to use the new platform. The other issue is the
continued splitting of content and esp. volunteers have to find extra time
to check those other platforms to stay in touch with the movement.

Be well and healthy

Matej

On Sunday, June 12, 2022, Galder Gonzalez Larrañaga 
wrote:

> Ceill,
> I am a big fan of having 'one front door' for people that are trying to
> find answers to questions. Having the front door in another building, with
> another technology, and once they are in we say them that our building is
> the other one, the one that is falling down (but don't visit the basement,
> please, is full of money) is the worst of the strategies.
>
> Best,
> Galder
> --
> *From:* Ciell Wikipedia 
> *Sent:* Sunday, June 12, 2022 6:03 PM
> *To:* Wikimedia Mailing List 
> *Subject:* [Wikimedia-l] Re: Join the new Movement Strategy Forum
> community review
>
> Hi,
>
> First: I am a big fan of having 'one front door' for people that are
> trying to find answers to questions they do not know where to ask (last
> year's movement communications insights on this
> ).
> I think a forum, actively moderated by people helping and pointing users to
> the right places, would be a huge improvement for community questions and
> input. Especially the one-click translation service is imho a big plus in
> service in comparison to Meta.
>
> It does however worry me that when I joined the forum last weekend to take
> a peek, I stumbled on a thread with a very specific question about Commons
> and giving permission via VRT. The thread had multiple replies, but no one
> had a real substantial answer. Well, replies were along the lines of 'No,
> there is no template for this' and 'This should be discussed on Commons'.
> While the answers were somewhat correct, they were obviously not helpful
> for the person asking this specific question and, as far as I could tell,
> none of the respondents were a member of our VRT teams. So this user was
> effectively not helped by posting the question on the forum.
> Even more so, because the question on the forum was not noticed by VRT
> agents (most of us working on the permissions queues and Commons will have
> the /Noticeboard on Commons on our watch list and can be pinged if country
> or language specific knowledge or advise is needed for a question), and
> secondly it will be more difficult for the people working from our end that
> will have to follow up if the person does decide to bring the question to
> Commons or VRT after all.
>
> Besides that, with my MCDC hat on, I hope after this trial period we'll
> get to see the data on how many people interacted about the Movement
> Strategy that we have not heard from in the previous 5 years through any of
> the other platforms that are in use to gather feedback. Already trying to
> watch several channels with Strategy discussions, I count on the MSG team
> to bring back these numbers and a summary of what is being discussed on the
> forum back to Meta. Even in a virtual world there is a limit on how many
> channels a Wikimedian can watch.
>
>
> *NB: I see Sj's response crossed mine while I was writing, but let my
> example underline the issue of 'no unified notifications' and a possible
> problem with 'coherent archiving'. *
>
> *Please also be aware G-translate does not know all languages we have
> projects in, some of which are however supported by Yandex that is an
> option to choose for the Wikipedia article translation tool already. *
>
> Best,
> Ciell
>
___
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/JF2P7S53WPYY7CY7JTFNBN5ZHOWLLDEO/
To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org

[Wikimedia-l] Re: Bing-ChatGPT

2023-03-18 Thread Matej Grochal
Hi all

I agree that the AI creators should attribute Wikipedia as their source.
But on the other hand, when the result is incorrect etc we might actually
be glad that they do not attribute it.
The issue is how to convince readers to come to the source (our projects)
rather than using in-between steps (AI or otherwise).

Matej

On Sat, Mar 18, 2023 at 3:22 AM Steven Walling 
wrote:

> On Fri, Mar 17, 2023 at 6:03 PM The Cunctator  wrote:
>
>> I really feel like we're getting into pretty aggressive corporate abuse
>> of the Wikipedia copyleft.
>>
>
> I completely agree. It makes me pretty angry that Wikipedians have spent
> millions of volunteer hours creating content to educate and inform people
> as accurately as we can, and it's being used to generate convincing but
> often wildly misleading bullshit.
>
> The ground truth on what generated AI content is (from a copyright
> position) and where authorship/ownership lies seems to be rapidly evolving.
> The U.S. Copyright Office recently refused to issue copyrights for some
> AI-generated works, seemingly on the principle that they lack human
> authorship / are essential to contracting work for hire from an artist or
> writer.
>
> IANAL of course, but to me this implies that responsibility for the
> *egregious* lack of attribution in models that rely substantially on
> Wikipedia is violating the Attribution requirements of CC licenses. Just
> like the Foundation took a principled position in testing the legality of
> warrantless mass surveillance, I would love to see us push back on the
> notion that it's legal or moral for OpenAI or any of these other companies
> to take our content and use it to flood the Internet with machine-generated
> word diarrhea.
>
>
>> On Fri, Mar 17, 2023, 4:45 PM Adam Sobieski 
>> wrote:
>>
>>> Hello,
>>>
>>> I would like to indicate "Copilot" in the Edge browser as being
>>> potentially relevant to Wikipedia [1][2].
>>>
>>> It is foreseeable that end-users will be able to open sidebars in their
>>> Web browsers and subsequently chat with large language models about the
>>> contents of specific Web documents, e.g., encyclopedia articles. Using Web
>>> browsers, there can be task contexts available, including the documents or
>>> articles in users' current tabs, potentially including users' scroll
>>> positions, potentially including users' selections or highlightings of
>>> content.
>>>
>>> I, for one, am thinking about how Web standards, e.g., Web schema, can
>>> be of use for amplifying these features and capabilities for end-users.
>>>
>>>
>>> Best regards,
>>> Adam Sobieski
>>>
>>> [1]
>>> https://learn.microsoft.com/en-us/deployedge/microsoft-edge-relnote-stable-channel?ranMID=24542#version-1110166141-march-13-2023
>>> [2] https://www.engadget.com/microsoft-edge-ai-copilot-184033427.html
>>>
>>> --
>>> *From:* Kimmo Virtanen 
>>> *Sent:* Friday, March 17, 2023 8:17 AM
>>> *To:* Wikimedia Mailing List 
>>> *Subject:* [Wikimedia-l] Re: Bing-ChatGPT
>>>
>>> Hi,
>>>
>>> The development of open-source large language models is going forward.
>>> The GPT-4 was released and it seems that it passed the Bar exam and tried
>>> to hire humans to solve catchpas which were too complex. However, the
>>> development in the open source and hacking side has been pretty fast and it
>>> seems that there are all the pieces for running LLM models in personal
>>> hardware (and in web browsers). Biggest missing piece is fine tuning of
>>> open source models such as Neox for the English language. For multilingual
>>> and multimodal (for example images+text) the model is also needed.
>>>
>>>
>>> So this is kind of a link dump for relevant things for creation of open
>>> source LLM model and service and also recap where the hacker community is
>>> now.
>>>
>>>
>>> 1.) Creation of an initial unaligned model.
>>>
>>>- Possible models
>>>   - 20b Neo(X)  by
>>>   EleutherAI (Apache 2.0)
>>>   - Fairseq Dense
>>>    by Facebook
>>>   (MIT-licence)
>>>   - LLaMa
>>>    by
>>>   Facebook (custom license, leaked research use only)
>>>   - Bloom  by Bigscience 
>>> (custom
>>>   license . open,
>>>   non-commercial)
>>>
>>>
>>> 2.) Fine-tuning or align
>>>
>>>- Example: Standford Alpaca is ChatGPT fine-tuned LLaMa
>>>   - Alpaca: A Strong, Replicable Instruction-Following Model
>>>   
>>>   - Train and run Stanford Alpaca on your own machine
>>>   
>>>   - Github: Alpaca-LoRA: Low-Rank LLaMA Instruct-Tuning
>>>   
>>>
>>>
>>> 3.) 8,4,3 bit-quantization of model for reduced hardware 

[Wikimedia-l] Wikimedians of Slovakia 2023 annual activity report

2024-05-21 Thread Matej Grochal
Dear community members

you might enjoy reading the Wikimedians of Slovakia 2023 Activity Report
<https://meta.wikimedia.org/wiki/Wikimedians_of_Slovakia/Reports/2023/en>.
With best wishes

Matej Grochal
WMSK
___
Wikimedia-l mailing list -- wikimedia-l@lists.wikimedia.org, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at 
https://lists.wikimedia.org/hyperkitty/list/wikimedia-l@lists.wikimedia.org/message/EOSXRMX3TAHV3NZ4ZSYKKALKNOVPB6DB/
To unsubscribe send an email to wikimedia-l-le...@lists.wikimedia.org