User-facing changelog for 4.1

[krassowski]

References

Closes #15576

For review: https://jupyterlab--15648.org.readthedocs.build/en/15648/getting_started/changelog.html#v4-1

• After approval but before merging the links to images need substituting krassowski/jupyterlab/changelog-4.1 → jupyterlab/jupyterlab/main

Code changes

None

User-facing changes

Changelog for upcoming 4.1 has high-level overview of the highlights

Backwards-incompatible changes

None

[jupyterlab-probot]

Thanks for making a pull request to jupyterlab!
To try out this branch on binder, follow this link:

[jtpio]

Thanks!

Looking good, left a couple of inline suggestions.

[jtpio]

Maybe we can leave a note to encourage users to open a post on Discourse if they have trouble updating?

[krassowski]

I agree in principle but rather than recommending opening a post, I would prefer if we directed to a specific post. It would be a bit of a moderation nightmare if many users started many similar topics. If you would like to create a post to start the discussion feel feel to add a link to it here.

[jtpio]

We can also just leave it like this and see if users open issues or not.

[jtpio]

cc @JasonWeill who may also be interested in reviewing the user facing changelog

[jtpio]

Thanks, looks good!

We can always iterate more on it until the final 4.1 release if needed.

But for now it would be good to have this user facing changelog available soon so we can more publicly invite users to try the beta.

[krassowski]

Thank you, I agree.

@JasonWeill I saw that you were back online and just wanted to check if you were already looking at this (no expectations, just wanted to make sure we will not merge if you are mid-flight with the review or had it planned for later today) - otherwise I will substitute the URLs in images and merge now, and we can iterate in follow-up PRs.

[JasonWeill]

@krassowski @jtpio Thanks for your patience! I've been out Mon–Wed this week on vacation, but I can look at this as soon as I can.

[JasonWeill]

Suggested change

	enabling multi-line completions such as those generated by large language models (like GPT-based copilots).
	allowing generative AI models to provide multi-line completions.

OpenAI asserts ownership of the name "GPT", and "CoPilot" is a trademark of Microsoft.

[krassowski]

But copilot lowercase might be ok? Do you have a source on GPT being a trademark?

[krassowski]

It looks like Microsoft applied for "MICROSOFT COPILOT" but has not been granted it (https://news.ycombinator.com/item?id=38823651) yet. I doubt they can just copyright a word used from 1920s ;) This is why in a previous iteration I suggested "GPT-based copilots" over "GitHub Copilot" suggested by Jeremy)

[JasonWeill]

"Copilot" has been used for ~100 years in aviation, but in generative AI terms, it seems vendor-specific. https://openai.com/brand says, 'The "OpenAI" name, the OpenAI logo, the "ChatGPT" and “GPT” brands, and other OpenAI trademarks, are property of OpenAI.' OpenAI's GPT trademark application was filed on December 27, 2022, and is "under examination" today. I'm biased as an employee of a company that makes large language models, but I don't think any one company's LLM-related intellectual property (irrespective of trademark approval status) should be in Jupyter's docs.

[jtpio]

This is why in a previous iteration I suggested "GPT-based copilots" over "GitHub Copilot" suggested by Jeremy)

Yes, the original suggestion was to have the word "copilot" in the changelog, because this is what most users have heard of, and also what they used when requesting the feature: #14267

The high number of 👍 suggests folks come to the repo and search for "copilot":

[krassowski]

Thanks for the reference on "GPT" trademark claim from OpenAI. It is interesting to see if their application will be accepted, especially because I would argue this is a generic term for a family of models. For anyone interested further, there is a section in Wiki on this: https://en.wikipedia.org/wiki/Generative_pre-trained_transformer#Brand_issues

Relatedly, OpenAI has applied to the United States Patent and Trademark Office (USPTO) to seek domestic trademark registration for the term “GPT” in the field of AI.[71] OpenAI sought to expedite handling of its application, but the USPTO declined that request in April 2023.[75] In May 2023, the USPTO responded to the application with a determination that "GPT" was both descriptive and generic.[76] As of November 2023, OpenAI continues to pursue its argument through the available processes. Regardless, failure to obtain a registered U.S. trademark does not preclude some level of common-law trademark rights in the U.S.,[77] and/or trademark rights in other countries.[78]

For any given type or scope of trademark protection in the U.S., OpenAI would need to establish that the term is actually “distinctive” to their specific offerings in addition to being a broader technical term for the kind of technology. Some media reports suggested that OpenAI may be able to obtain trademark registration based indirectly on the fame of its GPT-based chatbot product, ChatGPT,[75][79] for which OpenAI has separately sought protection (and which it has sought to enforce more strongly).[80] Other reports have indicated that registration for the bare term “GPT” seems unlikely to be granted,[71][81] as it is used frequently as a common term to refer simply to AI systems that involve generative pre-trained transformers.[3][82][83][84] In any event, to whatever extent exclusive rights in the term may occur the U.S., others would need to avoid using it for similar products or services in ways likely to cause confusion.[81][85] If such rights ever became broad enough to implicate other well-established uses in the field, the trademark doctrine of descriptive fair use could still preserve some room to continue non-brand-related usage.[86]

But yeah, happy to not include this in here if this raises an objection :)

I'm biased as an employee of a company that makes large language models, but I don't think any one company's LLM-related intellectual property (irrespective of trademark approval status) should be in Jupyter's docs.

I believe we agree on this point, the question is whether this particular term represents "any one company's LLM-related intellectual property" or just a generic term as introduced to scholarly community by scholarly means and simultaneously developed into a commercial product (ChatGPT).

[JasonWeill]

Avoids passive voice, per writing style guide

[fcollonval]

Thanks a lot @krassowski

It looks great. I was wondering if we should mentioned the change in shortcut capture. The migration guide may be a better fit, though.