solomatov

9 ms·

91.
by solomatov 3y ago
It’s very interesting how enforceable this trademark policy is. There’re trademark fair use, and it looks like they try to take away some part of it, through forbidding to use word Rust in other product names. P.S. I am not a lawyer.
92.
by solomatov 3y ago
Policies aren't always binding. There're companies which include a point about complying with the trademark policies in their terms of use, there're some which just located on their site under the name guidance. P.S. I am not
93.
by solomatov 3y ago
Which is completely unsurprising since ROS (Reactive Oxygen Species) are used by organisms to kill cancer cells, bacteria and other foreign stuff in the body. P.S. I am neither a doctor nor a biologist.
94.
by solomatov 3y ago
I don't exactly understand how you could use it for commercial use in the original license quoted in the thread: >You may download or copy the Content (and other items displayed on the Services for download) for personal non-commerc
95.
by solomatov 3y ago
Are you aware of any easy to install distros with such kernels?
96.
by solomatov 3y ago
Is anyone aware of plans to port Asahi changes to make other linux distros, i.e. Ubuntu, and others, to be able to run without additional setup on M1 Macs?
97.
by solomatov 3y ago
I would call understanding what's the underlying mechanism behind LLMs and other deep models such a problem. I.e. what's the internal representation? How it emerges over training? What its failure modes. There're two reasons:
98.
by solomatov 3y ago
>Even if this does prove effective in humans, eventually cancer cell defenses will evolve to get around it. Cancer becomes tolerant to drugs separately for each patient. I.e. initially it isn't resistant, but the drug used on a pati
99.
by solomatov 3y ago
It's actually possible to do this. There's a thing which is called perpetual bonds, i.e. bonds which pay a fixed sum ad infinitum. Not surprisingly, they have a non infinite cost. ( https://en.wikipedia.org/wiki&#x2
100.
by solomatov 3y ago
Yep, I agree with you that it's better to be explicit than implicit here. Initialization/destruction order is often tricky, and source of obscure bugs.
101.
by solomatov 3y ago
> No destructors. What's the logic behind this? It's nice to have a way to pair resource usage with disposal. Or does the linear type system allows not to forget about the resources which have to be closed.
102.
by solomatov 3y ago
Smalltalk has this "feature" and was doing fine. It was pretty popular at some point.
103.
by solomatov 3y ago
So as I see it, it's possible to teach SOTA LLM with prompts to emulate a state machine part of the Turing machine. Am I right?
104.
by solomatov 3y ago
Ok. I see. Sorry.
105.
by solomatov 3y ago
> Are large language models even Turing complete? If you want to see where they have problem ask them to do something about deep hierarchical objects. For example, consider this prompt: "Draw me a complete binary tree with numbers f
106.
by solomatov 3y ago
You are mixing up LLMs with Transformers. Transformers with memory are Turing complete, but AFAIK, current state of the art LLMs aren't trained with any kind of memory.
107.
by solomatov 3y ago
It's possible to argue that there's no such a thing as infinite precision floats. I.e. all objects have limited "size", and real numbers is just a nice abstraction to simplify discrete world. ( https://en.wikip
108.
by solomatov 3y ago
It's not realistically Turing complete. It assumes infinite precision.
109.
by solomatov 3y ago
IMO, it's a replacement for C/C++ with higher safety guarantees. Of course it's more suitable for ML, and Web, than C/C++, but if there're no hard perf requirements, there're better options.
110.
by solomatov 3y ago
Systems where you need to prove correctness at least informally. It’s much easier to prove correctness of a functional program than imperative one.
111.
by solomatov 3y ago
Something with intention to make you money and directly related to powerful machine, i.e. writing code, editing videos, images, etc. As opposed to casual use where you plan simple games, browse internet, etc.
112.
by solomatov 3y ago
> would make AR coding more productive than simply on a monitor One obvious thing is much more space. I.e. unlimited number of extra monitors, or entities which work as such.
113.
by solomatov 3y ago
Not always, if you are connected to a source of electricity and doing something limited by speed performance is critical. For me, Desktop use is almost perfect on Apple due to battery life and perf but professional use is much better on Int
114.
by solomatov 3y ago
My understanding is that the book is based on this PhD dissertation.
115.
by solomatov 3y ago
There's a book which looks like it's based on this dissertation, which probably is a better source to read if you are interested in this topic: https://www.amazon.com/Purely-Functional-Data-Structures-Oka...
116.
by solomatov 3y ago
> Between the apparently-probabilistic nature of LLMs deciding which instructions ought to be followed It's not that probabilistic if you want it to be. When sampling from LLMs, you put a temperature parameter, and if it's 0, i
117.
by solomatov 4y ago
It's not that simple. You couldn't just use any random GPU in the cloud, you need to connect them using the right kind of interconnect (Mellanox bough by NVidia) which is very expensive.
118.
by solomatov 4y ago
>In a completely manually managed language like C or C++, you can handle fragmentation problems yourself by writing your own allocators or doing object pools to reuse previously allocated memory. You have control over fragmentation. It&#
119.
by solomatov 4y ago
IMO, the problem is that it doesn't work. It could introduce other bugs, make incorrect changes, or mess up the code. If a good developer does this, it converges, with GPT I am not so sure. We need to find a way for the system to itera
120.
by solomatov 4y ago
It's definitely useful, but IMO it's not 5x or 10x and of course not x1000 force multiplier. More like a very good and useful tool.

More ›