Talk:North American Hub Research Project

Main

Talk

Reading materials

Research and findings

Report and implementation plan

Welcome!

We welcome your questions and input. - Ariel Cetrone (WMDC) (talk) 15:12, 11 May 2023 (UTC)

Hi! In case that advisors to the advisory committee are welcome/needed, the CEE Hub would be happy to support all of you in this endeavour :-) Cheers, Philip Kopetzky (talk) 20:57, 18 May 2023 (UTC)

Hello, @Philip Kopetzky, and thank you! It would help us to meet with you. We have been learning from the CEE hub's experience, and from other affiliates. I'll email you. -- econterms (talk) 00:54, 24 July 2023 (UTC)

Good use of time and money

There are so many urgent needs that the Wikimedia Foundation could be working on but isn't. Copyright backlogs on Commons, to a lesser extent on enwiki,, and other wikis. Draft backlogs on AfC and NPP on English Wikipedia. The AfC backlog is often months long which is not acceptable given that it's the first exposure of a new contributor to Wikipedia. But it's pretty hard to deal with since volunteers don't want to wade through spam all day. I think it would be a better use of money to pay people to proactively patrol AfC and NPP, or develop automated copyvio detection for Commons, than half of the things that the foundation spends money on. Buidhe (talk) 00:21, 12 November 2023 (UTC)

A conversation with ChatGPT

Buaidh: Does ChatGPT use Wikipedia as a source?

ChatGPT: Yes, ChatGPT has been trained on a mixture of licensed data, data created by human trainers, and publicly available data. This corpus includes text from a wide range of sources, and Wikipedia is one of the sources used during the training process. The model is not, however, provided with direct access to databases, proprietary websites, classified information, unpublished works, or confidential sources. The information from Wikipedia and other sources is used to help the model generate responses to a variety of user inputs.

It's important to note that while Wikipedia is a valuable source of information, it can have inaccuracies or reflect different perspectives. Users should always verify critical information from multiple sources when necessary, especially for important or sensitive topics.

Buaidh: Why does ChatGPT use sources that have been copyrighted?

ChatGPT: ChatGPT is trained on a mixture of licensed data, data created by human trainers, and publicly available data. This includes text from a wide range of sources, and some of the sources may be copyrighted. The use of copyrighted material is a part of the training process to expose the model to a diverse range of language patterns and information available on the internet.

It's important to clarify that the model itself doesn't have access to specific copyrighted texts or databases. Instead, it learns patterns and associations from the data it was trained on, and it uses that knowledge to generate responses based on user inputs.

OpenAI, the organization behind ChatGPT, respects copyright and intellectual property rights. The use of copyrighted material is in accordance with fair use policies for research, education, and training purposes. Users of ChatGPT should be aware that while the model can provide information and generate text on a variety of topics, it's not a substitute for professional advice or information obtained from authoritative and reliable sources. If copyright concerns arise, it's recommended to consult legal advice or adhere to copyright policies.

 Buaidh  01:40, 29 December 2023 (UTC)

How does your post relate to the Hub project? OhanaUnited^{Talk page} 03:08, 29 December 2023 (UTC)

Wikimedia must contend with the development of artificial intelligence, and the NAhub can assist. Yours aye,  Buaidh  02:30, 30 December 2023 (UTC)

A trilingual North American Hub

A North American Hub should be trilingual for users of English, Spanish (Wikimedia América del Norte?), and French (Wikimédia Amérique du Nord?). Real-time translation and transcription software would promote communications among speakers of these three languages as well as the hearing impaired. While real-time translation and transcription software has not always been accurate, it is improving (often with artificial intelligence) and clarification questions can be raised promptly to resolve misunderstandings.  Buaidh  03:28, 22 May 2024 (UTC)

Central America

Although Central America currently has no Wikimedia user groups, it should be included in a North American Hub. Like the Caribbean, Central America is a geographic part of North America. Inclusion in a North American Hub will encourage regional participation in the Wikimedia movement.  Buaidh  01:08, 1 July 2024 (UTC)

Updates?

@Econterms: could someone share an update from WCNA this year?

This seems like a good point to highlight concerns and address them; there will be a long tail of concerns that touches a majority of potential members. And to update how other hubs are developing.

From my perspective, I'd like to see early focus on two things:

• Support for WCNA, working with the user group + hosts
• Support for communication and peer coordination among smaller groups, outside of annual events.

And making sure that the group discussing how to proceed includes people from a range of regions and participation styles. –SJ talk  19:14, 9 October 2024 (UTC)