LightMachine

6 ms·

91.
by LightMachine 3y ago
100% agree - we're working on exactly that for the upcoming release in a few months!
92.
by LightMachine 3y ago
Good question! There are several implementations of the full λ-calculus on interaction nets; that isn't the problem. The problem is that these solutions always came up with a constant time slowdown, which made interaction nets even les
93.
HVM updates: simplifications, finally runs on GPUs, 80x speedup on RTX 4090(twitter.com)
91 points by LightMachine 3y ago|29 comments
94.
by LightMachine 3y ago
As someone who has been studying type theory and developing strongly typed programming languages, that is a very strange take to me. It isn't untyped, it has a type: the GeometricAlgebra type, which just happens to have a very rich str
95.
by LightMachine 3y ago
Watched and read it all and find it amazing work. Text editing on mobile is embarassingly terrible and the proposals presented are very solid. Props to the author.
96.
by LightMachine 3y ago
Shut up and accept what? A company being closed source / for profit? Are we rioting against capitalism or something? Should I get my pitchforks? Just to be sure, we'll include literally all major tech companies, or is only OpenAI
97.
by LightMachine 3y ago
We get it. OpenAI is a for profit company now. The "Open" in name doesn't reflect their vision anymore. Can we stop it already? Or is this some kind of rant against for profit / closed source in general?
98.
A Lock-Free Interaction Combinator Evaluator to Fix LoL' Client Performance(twitter.com)
2 points by LightMachine 3y ago|0 comments
99.
by LightMachine 3y ago
why there are no good videos of superfluids in action in the whole internet?
100.
by LightMachine 3y ago
GPT-4 does it very well though. (Not saying OP used it, was just curious if that'd work.) Prompt: > Observation of zero resistance above 100∘ K in Pb10−xCux(PO4)6O > convert this to a title, with the correct usage of unicode Outp
101.
wrong url :((twitter.com)
1 points by LightMachine 3y ago|0 comments
102.
by LightMachine 3y ago
I'm Brazilian and Pix is great. No reason to use anything else. Our government is terrible, but our central bank is actually quite competent. Credit to where it's due.
103.
by LightMachine 3y ago
spent 2 minutes on captcha. giving up. fts
104.
by LightMachine 3y ago
I'm in São Paulo right now and I didn't even notice that. I guess everywhere else is just flooded with billboards then?
105.
by LightMachine 3y ago
Well, then perhaps you need to reflect where your research failed, because it clearly did. Everyone is using Discord; from adult professionals to clueless kids who barely know how to connect to Wifi. So, if all these clueless kids are there
106.
Show HN: ChatSH – a terminal where GPT writes the commands for you(github.com)
7 points by LightMachine 3y ago|0 comments
107.
by LightMachine 3y ago
Any idea how much it cost to train it and how it was trained?
108.
Implementing complex numbers and FFT with just algebraic datatypes (no floats)(gist.github.com)
8 points by LightMachine 3y ago|0 comments
109.
by LightMachine 3y ago
Where have you read that HVM beats C or Rust? That is not something I've ever written and we're far from that, specially considering the budget that has been thrown into Rust, although we could (and hope to) get there one day. Thi
110.
by LightMachine 3y ago
(Also, exponentially faster is used on the technical sense here! Optimal evaluators can reduce some programs with linear time complexity, while conventional runtimes are exponential.)
111.
by LightMachine 3y ago
Well, it should be a viable alternative to GHC once we implement full lambdas, but without them, it should be seen as a separate language, just like Rust. The complete subset of lambda terms that HVM can soundly evaluate hasn't been id
112.
by LightMachine 3y ago
I think this kind of sentence makes such a huge disservice for the HVM. Rust closures also doesn't support the full λ-calculus! HVM supports way more λ-terms than Rust does. That said, we're working on full λ-calculus support, s
113.
by LightMachine 3y ago
Very well written and beautiful article.
114.
by LightMachine 3y ago
I feel like no less than 10 years if the singularity doesn't kick in before that. Hardware and energy isn't progressing as fast as we'd like, and that is the main bottleneck. As in, imagine a world where we actually had the s
115.
by LightMachine 3y ago
I agree with most of your points, we should definitely rewrite and tone down the README. The FAQ was written a few months after, due to that kind of negative reception. To address your points, note that by "superseding" the Turing
116.
by LightMachine 3y ago
Inpla is a very interesting project that is more elegant than HVM in several regards (as it doesn't need to worry about functional programming!) and should be more known. I've added a link to it from the HVM's repo a few mont
117.
by LightMachine 3y ago
Eventually, definitely, but less likely for this round.
118.
by LightMachine 3y ago
I love WASM - who doesn't? Maintaining a WASM target was included on HOC's current fundraise, and is essential to our JIT plans. I don't know much about typed assembly, but I guess it would enable optimizations that aren&#x27
119.
by LightMachine 3y ago
Yes! But first we want to implement some essential features, like full lambda support and some high priority optimizations. Once HVM is ready to support production software, we'll provide docs on how to best use it as a general compila
120.
by LightMachine 3y ago
Absal is just an early implementation and experiment. It is written in JS and not meant to be a production-ready runtime and general purpose compilation target. HVM has that goal. That said, Absal is a great repo to study and learn the theo

More ›