mlajtos

8 ms·

61.
by mlajtos 2y ago
Yes, the handwriting recognition system needs to communicate results as soon as possible to instill trust in the user. I need to be able to spot and fix misinterpretations as quickly as possible, otherwise errors accumulate real fast.
62.
by mlajtos 2y ago
I thought the same few years ago – putting recognized symbol above is simple and effective. But you can get the same effect (confidence in recognized input) with the Smart Script technique when done right. I briefly experimented with that i
63.
by mlajtos 2y ago
Fair point. Interactivity is a spectrum that should match the user preference. Interactivity helps when you are learning the tool, because you can do & fix mistakes faster. But when you want to think deeply, it should stand away and be
64.
by mlajtos 2y ago
In the article I discuss tech that doesn't exist yet.
65.
by mlajtos 2y ago
Fixed, thank you for your suggestion! (I love that quote.)
66.
by mlajtos 2y ago
You can do this in Math Notes – scribble numbers in a column, underline it and boom, there is your result. Doesn't work in typed text though.
67.
by mlajtos 2y ago
Yes, having an ability to do simple math across the whole ecosystem is extremely useful.
68.
My Notes on Apple Math Notes(mlajtos.mu)
178 points by mlajtos 2y ago|67 comments
69.
by mlajtos 2y ago
Usually, people use slide rule and handheld calculator in combination with paper and pencil. Combo of writing tool & computing tool. Merging them is a good idea.
70.
by mlajtos 2y ago
With Math Notes, you just write with your hand `log₂(1024)=` and you'll see the result. No slide rules, no buttons, just magic paper that does math. In this case, the logarithm is the tool I want – not slide rule, not handheld calcul
71.
by mlajtos 2y ago
Computers are faster at lookup tables than humans.
72.
by mlajtos 2y ago
Parsing traditional 2D math notation is super hard problem and this is first public version. Things will get better with time. But you hit the nail on the head with "narrow undefined subset". Documentation on Math Notes is almost
73.
by mlajtos 2y ago
iPhone did less than competitors. Ultimately, Math Notes will get copied and sold for less.
74.
by mlajtos 2y ago
Ironic how one demonstrates a questionable grasp on market dynamics if they think a feature of an expensive iDevice won't be copied by competitors and sold for less.
75.
by mlajtos 2y ago
But better calculator does revolutionize math education. Math is about problem-solving – if I can use tool that speeds up the boring parts (those that can be automated), I can focus on what really matters. For example, doing logarithms wi
76.
by mlajtos 2y ago
iPad mini 2024 with Pencil Pro and Math Notes is going to revolutionize math education.
77.
by mlajtos 2y ago
That is pretty cool. I found a follow up work that applies this technique to LLM: https://konczer.github.io/doc/Poster_EEML23_JozsefKonczer.pd...
78.
by mlajtos 2y ago
Can you tell me how it compares to Deepnote.com please?
79.
by mlajtos 2y ago
It was research project. Those rarely get out to public.
80.
by mlajtos 2y ago
Inkbase is pretty far from Math Notes from iPadOS 18. Superficially they are both "programming with pencil". Math Notes are more about numerical computing – numbers, symbols, quantities, math, etc. Inkbase tackles different proble
81.
by mlajtos 2y ago
Have you seen Apple Math Notes? Impressive first version with great promise for the future.
82.
by mlajtos 2y ago
If you are into this kind of thing, you might like these essays from 2021: 1. https://mlajtos.mu/posts/new-kind-of-paper 2. https://mlajtos.mu/posts/new-kind-of-paper-2 3. https://mlajt
83.
by mlajtos 2y ago
iPadOS 18 is in public beta right now. Will be out in September.
84.
How to Give Insanely Great Apple Vision Pro Demo(mlajtos.mu)
2 points by mlajtos 2y ago|0 comments
85.
by mlajtos 3y ago
Exactly! (Until the lens cups your whole eye socket and delivers your natural FOV.)
86.
by mlajtos 3y ago
I am Apple customer and owned multiple VR headsets since 2016. Stepping outside the Apple ecosystem is always painful. They are definitely scratching my itch.
87.
by mlajtos 3y ago
Neural nets usually use 32-bit floats for weights. There is a rapid trend of using quantization. There are already papers describing 2-bit quantization strategies – either quantization-aware training or post-training quantization. Less prec
88.
by mlajtos 3y ago
I am big proponent of balanced ternary for neural nets and general computation (Setun), but this repo does not make sense to me at all.
89.
by mlajtos 3y ago
For JavaScript it is ES1995. https://github.com/mlajtos/es1995
90.
by mlajtos 3y ago
N9 was a work of art. Fuck Elop.

More ›