Announced: Ainote 2

[DrChiper]

I just stumbled on a kickstarter announcement for "The World's Thinnest AI-Powered Paper Tablet". See here It supports ereading too.
Prices comparable with Remarkable Move 2, but seemingly you get more bang for the buck$. As it is not there yet, the reality might prove otherwise.

[rantanplan]

Why do people want AI slop on their handwritten note device?

[DrChiper]

Originally Posted by rantanplan

Why do people want AI slop on their handwritten note device?

Why would people need a handwritten note DEVICE where just paper would suffice?

[Quoth]

Originally Posted by rantanplan

Why do people want AI slop on their handwritten note device?

Probably they don't. Of course there is AI (= stuff they think sells better if labelled AI), AI (as in LLM and so called generative AI) and even AI (Actual Indians).

[salamanderjuice]

Originally Posted by Quoth

Probably they don't. Of course there is AI (= stuff they think sells better if labelled AI), AI (as in LLM and so called generative AI) and even AI (Actual Indians).

One of the few things AI is very very good at is summarizing large quantities of text. I could see that working from a notetaking perspective. Take course notes or something, have the LLM create a summary of what's important from the 200 pages of notes.

[Cactus Chef]

There are occasions where LLMs could be useful.

See:
https://www.youtube.com/watch?v=u6qxixgQJ4M

[nana77]

Maybe it will help AWS AIs on futue handwritings' developments .

[Quoth]

Originally Posted by salamanderjuice

One of the few things AI is very very good at is summarizing large quantities of text. I could see that working from a notetaking perspective. Take course notes or something, have the LLM create a summary of what's important from the 200 pages of notes.

It's been proven to be unreliable. See BBC vs Apple or problems with board meetings. Of Google search summaries.

That's simply LLM AI company propaganda. Randomly unreliable summaries are useless. They have to be checked and that takes as long as doing them.

[Quoth]

Originally Posted by Cactus Chef

There are occasions where LLMs could be useful.

As a toy, or amusement. There is an incredible amount of propaganda sponsored by the corporations.
e.g. https://arstechnica.com/ai/2025/09/e..._medium=social

There are useful AI applications, at least they are marketed as AI. But they are not LLM models or so-called generative AI.

[Graham44]

Great googly moogly something else they've managed to shoehorn AI into; I'm beginning to think AI really stands for Actual Insanity.

Looking at the Kickstarter it's using ChatGPT an AI that one study found got answers wrong 52% of the time - https://dl.acm.org/doi/10.1145/3613904.3642596 , and other research found it got the answers wrong 58% of the time (going by the charts) - https://www.cjr.org/tow_center/we-co...iting-news.php

I mean surely if you are going to go to the trouble of taking the notes down in the first place you could summarize them yourself? Instead of relying on the hallucinating insanity that is AI?

And they want you to pay over £300 at least for the privilege!

[tomsem]

It's very similar to this, and price is about the same:

https://viwoods.com/products/viwoods-aipaper

...which has been out for about a year now and has had a number of of updates.

AI for analyzing medical imaging is proving very useful and is able to outperform humans significantly. More of this, please.

But otherwise, it's way over its skis, out of control, and leading to unknown and likely negative outcomes for us humans. Even the known downsides (slave labor to curate results and skyrocketing energy usage) are bad enough.

[Quoth]

Originally Posted by tomsem

AI for analyzing medical imaging is proving very useful and is able to outperform humans significantly.

Actually not as good as the sellers claim. It's actually pattern matching based on scanned of images classified by human experts. Indeed that kind of "AI" (not actually AI, but AI branding) can save time.

Nothing to do with the kind of AI claimed for summarising notes.

1. Don't take anything in the popular media at face value.
2. Check who paid for the press release or "paper".
3. Proper explanation and peer review.

The "Medical Watson" was just Watson branding. Not the same sw as won "Jeopardy". It was poor and the big USA hospital eventually sued IBM.