enricozb

5 ms·

31.
by enricozb 1y ago
When I say "how did we get here", I don't mean "how did we end up with these opinions (e.g. racism) on our soil". I mean something more like: 1. Why is discussing these things so difficult? So many internet forums a
32.
by enricozb 1y ago
I have such disdain for the e/acc crowd given that I believe that "we do not understand the consequences of what we are building". But now I'm not sure if it's fair to ignore the consequences of building Twitter, or
33.
by enricozb 1y ago
I doubt that people who bike their kids in baskets daily in Amsterdam would do the same in NYC, given how dangerous it is. The risk tolerance required to bike in NYC is much higher.
34.
by enricozb 1y ago
NYC is so dense it could be a bikers' paradise in the US. Why are we supporting even more car infrastructure :/
35.
by enricozb 1y ago
Hexagony is another 2d language: https://esolangs.org/wiki/Hexagony
36.
by enricozb 1y ago
I just remap my terminal to have Ctrl-X to send SIGHUP, and Ctrl-C to copy. X looks more like a kill to me anyways.
37.
by enricozb 1y ago
This article seems to argue that the cause of the inequality is the lack of taxes
38.
by enricozb 1y ago
However it also prevents hoarding wealth, which in turn prevents special interest groups and some forms of election manipulation. Balancing taxes for fairness and innovation is quite tricky...
39.
by enricozb 1y ago
As far as I understand, The lead of this project Kevin Buzzard is a mathematician first. And the majority of mathematicians are untroubled by non-constructive proofs. I would imagine that proof directions that result in the most interesting
40.
by enricozb 1y ago
Why doesn't react switch to doing a more sensible equality check, rather than relying on JS's unfortunate choices? (E.g. value equality for arrays and objects) I feel like this would simplify most existing sensible code.
41.
by enricozb 1y ago
Interaction nets are another computational model based on graph-reduction. https://ezb.io/thoughts/interaction_nets/lambda_calculus/202...
42.
by enricozb 1y ago
One benefit of tailwind is that while all of the classes look like they set properties with fixed values, they are actually almost all controlled by variables. This makes it easier to retain a coherent design during redesigns. Doing the sam
43.
by enricozb 1y ago
I'm not sure how these two things are related. When writing an `impl` for a struct, there's no assumption on the bounds of the generics (if any) unless they are specified _at the impl site_. For example, the bounds of T in i
44.
by enricozb 1y ago
For structs, why couldn't rust check the necessary bounds on `T` for each field to be cloned? E.g. in #[derive(Clone)] struct Weird<T> { ptr: Arc<T>, tup: (T, usize) } for `ptr`, `Arc<T>:
45.
by enricozb 1y ago
Fantastic. Been meaning to put something up like this myself. I feel like I gain so much information from just watching someone work, and having them open themselves up to questions.
46.
by enricozb 1y ago
Since I do not really use Ctrl-X on the terminal to cut text, I have changed my terminal configs to have Ctrl-X be sigint, and Ctrl-C to be copy. Makes for a nicer experience, also X looks more like a "kill" :)
47.
by enricozb 1y ago
Haven't looked at the link, but I think for a scenario like trading where there are market open and close times, you can just disable the GC, and restart the program after market close.
48.
by enricozb 1y ago
I'm not a mathematician (but I've seen exact sequences and commutative diagrams) and to me the stuff out of his IUT papers[0] looks borderline LLM-generated. I can only imagine what the LaTeX source looks like. [0]: https:/&
49.
by enricozb 1y ago
Thanks! The main thing that I had difficulty with, which you can see from the typst source, is the rendering and spacing around inline math symbols. They are rendered as SVGs, which is fine but then copying them doesn't work. Their spa
50.
by enricozb 1y ago
It works really well, I've written three blog posts using typst. Post: https://ezb.io/thoughts/interaction_nets/lambda_calculus/202... Typst source: https://github.com/enricozb/enric
51.
The Intelligence Curse(intelligence-curse.ai)
1 points by enricozb 1y ago|0 comments
52.
by enricozb 1y ago
Lambda variables are also scoped, in the sense that a bound variable shadows one with the same name if it is in scope at bind-time. When looking at lambda calculus as purely a term rewriting system, the rule for substitution is just a bit n
53.
by enricozb 1y ago
Differentiable Logic Gate Networks [0] are super interesting. However, I still don't like that the wiring is fixed initially rather than learned. I did some extremely rough research into doing learnable wirings [1], but couldn't g
54.
by enricozb 1y ago
Maybe Sciral Consistency?
55.
by enricozb 1y ago
nominative determinism
56.
by enricozb 1y ago
Yes, you're not the first to have asked this. I'll add one before I publish the next post.
57.
by enricozb 1y ago
Working on writing a series of blog posts on interaction nets. The first one is up already an is an introduction to the topic [0]. I feel like the topic is very accessible but the ideas are spread out across discord channels, papers, twitte
58.
by enricozb 2y ago
Working on ARC-AGI (2). Trying a few different approaches including gradient descent mechanisms to construct programs in DSLs. Was really inspired by the recent differentiable logic gate networks work from Google, and believe there is a way
59.
by enricozb 2y ago
Somewhat related: the Lambek calculus[0], which is kind of a general type system for written languages. Words are either functions or concrete types, and functions can take their arguments on the left or right.
60.
by enricozb 2y ago
I wish fd and rg would align some of their flags. For example, both fd and rg have a --type, but for fd it means file/directory/symlink etc. For rg it means the file MIME type. Another example is that fd has an --extension flag, a

More ›