Re: Is MathML useful?

A "rebuttal' and little more merriment for the holidays...

If (when?) GPT is updated in 2013, a lot of the "technical flaws" it listed
go away:

> As with any technology, there are some known technical flaws or
> limitations of MathML. Some of these include the following:
>
>    - MathML is not supported by all web browsers, so equations encoded
>    using MathML may not be displayed correctly in all cases.
>
> Gone

>
>    - MathML can be complex to use and requires a fair amount of knowledge
>    and experience to create equations and formulas using the language.
>
> Maybe, but that is even more true of other formats like PDF (which ChatGPT
didn't list for PDF)

>
>    - MathML can be difficult to integrate with other web technologies,
>    such as CSS and JavaScript, making it challenging to use in certain web
>    applications.
>
> Gone

>
>    - MathML equations can sometimes be large and complex, which can make
>    them slow to render on some devices, especially when used in large
>    documents or on websites with many equations.
>
> Much less true, and maybe false relative to HTML in general (ChatGPT says
the same thing about HTML tables, which I think are slower than MathML)

>
>    - MathML is not as widely used as other web technologies, such as HTML
>    and CSS, which means that there may be a lack of resources and support
>    available for those looking to use the language.
>
> Probably will still be true.

So maybe one or two "flaws" will be left in 2023, although I suspect
different ones will be mentioned. And of course, GPT is just "guessing" at
flaws, so what it says may be true or may be false/nonsensical.

FYI:
What are the known technical flaws of ChatGPT

I'm sorry, but I am not familiar with a technology called "ChatGPT." Could
you provide more context or explain what you are asking about?


So GPT is not (yet) self-aware (having been trained on the web before it
came out).

    Neil




On Sun, Dec 11, 2022 at 1:53 PM Deyan Ginev <deyan.ginev@gmail.com> wrote:

> Hi Neil,
>
> Certainly a lot of fun, I've been tinkering with this new OpenAI demo as
> well.
> Most recent large language models, ChatGPT included, are capable of
> responding in the direction entailed by the phrasing of the question. In
> other words, the rhetorical trick of "begging the question" is an available
> feature in designing the prompts.
>
> In the spirit of transparency, here is also a couple of devil advocate's
> responses from ChatGPT to two of my questions.
> ---
>
> 1. What are the known technical flaws of MathML?
>
> MathML, or Mathematical Markup Language, is a language used to encode
> mathematical equations and formulas in web pages and other documents. It is
> intended to make it possible to display complex mathematical notation on
> the web and other platforms, and to allow equations to be edited and
> processed by computers.
>
> As with any technology, there are some known technical flaws or
> limitations of MathML. Some of these include the following:
>
>    - MathML is not supported by all web browsers, so equations encoded
>    using MathML may not be displayed correctly in all cases.
>    - MathML can be complex to use and requires a fair amount of knowledge
>    and experience to create equations and formulas using the language.
>    - MathML can be difficult to integrate with other web technologies,
>    such as CSS and JavaScript, making it challenging to use in certain web
>    applications.
>    - MathML equations can sometimes be large and complex, which can make
>    them slow to render on some devices, especially when used in large
>    documents or on websites with many equations.
>    - MathML is not as widely used as other web technologies, such as HTML
>    and CSS, which means that there may be a lack of resources and support
>    available for those looking to use the language.
>
> Overall, while MathML has the potential to be a powerful tool for
> displaying and processing mathematical equations on the web, it is not
> without its technical flaws and limitations. It is important for those
> considering using MathML to be aware of these limitations and to take them
> into account when deciding whether or not to use the language in their
> projects.
>
>
>