The artist needs to use the brain and translate a lot, and I mean an awful lot, of the information seen in the model (3d space) into 2d.

Close one eye or put an eyepatch on. I’d expect this makes 3D -> 2D transformations easier after a while.

OpenShot. But that’s a video editor and not a conventional video player per se.

https://www.openshot.org/

media player that plays almost everything

What doesn’t it play?

There are alarm clock apps which can help. You may configure how unforgiving the alarm becomes.

https://play.google.com/store/apps/details?id=com.kog.alarmclock

Changed my life.

If we’re speaking of transformer models like ChatGPT, BERT or whatever: They don’t have memory at all.

The closest thing that resembles memory is the accepted length of the input sequence combined with the attention mechanism. (If left unmodified though, this will lead to a quadratic increase in computation time the longer that sequence becomes.) And since the attention weights are a learned property, it is in practise probable that earlier tokens of the input sequence get basically ignored the further they lie “in the past”, as they usually do not contribute much to the current context.

“In the past”: Transformers technically “see” the whole input sequence at once. But they are equipped with positional encoding which incorporates spatial and/or temporal ordering into the input sequence (e.g., position of words in a sentence). That way they can model sequential relationships as those found in natural language (sentences), videos, movement trajectories and other kinds of contextually coherent sequences.

There is a neat Firefox addon which makes visits to yubbtub more tolerable by getting rid of unrelated suggestions:
RYS — Remove YouTube Suggestions

It has a lot of options to tailor it to your needs.

*really

You can easily look them up using a search engine of your choice. But I understand the lazyness.