We used it to make memes (single sentence prompts) and the results are hilarious

How Jeb! Won the election: https://math-gpt.org/?video_id=b9e16dbd-cab6-44e3-a393-f255b...

Mathematical proof AI is doomed: https://math-gpt.org/?video_id=c5745f99-b1d1-4c57-8a93-6fe3b...

I tried it on a simple problem and generated the video thinking I would get a video that would go through the steps of division. Instead it said "Let's take number XX and divide it by number YY using a calculator to get the result ZZ".

I'm thinking, wait a tick, you're the calculator. Weren't you supposed to show your work? It's that literally the point?

It's a funny utility but I'm not sure it's all that useful.

FYI for the input "Graph a cartoid and explain it" I get a correct definition and some interesting facts/properties (in spite of the fact that I mis-remembered the name of the object I had in mind; it should have been "cardioid"). But then it tries to provide a plot, and the plot has an error:

``` Error Please check your input: Undefined variable heta ```

Looks like there's a "theta" that's being incorrectly truncated, or something to that effect.

I refreshed a few times and found that I consistently saw this same error.

That's a really interesting application of LLMs and manim.

My problem with it, as with LLMs in general, is that i can't fully trust it. This isn't a problem with topics you know well. But those 3b1b style tutorials are meant to build an understanding or mental model of a concept for people who don't know it well yet. I'm afraid that, if the output is (partially) wrong, watching such a video could be even harmful in understanding the concept, because your 'greenfield' brain may pick up pieces of it that made sense to it.

Sorry, this one is really terrible (also I asked for no audio) https://math-gpt.org/?video_id=6f622c5b-ccf3-408f-9db1-56d63...