HN user

mapehe

209 karma
Posts34
Comments38
View on HN
bytesauna.com 6mo ago

Where 2025's agentic AI hype fell short

mapehe
1pts1
bytesauna.com 6mo ago

Language Is the New UI

mapehe
2pts1
bytesauna.com 6mo ago

Insight Debt

mapehe
1pts1
bytesauna.com 7mo ago

What Is Creativity?

mapehe
1pts1
bytesauna.com 7mo ago

Understanding Mathematics Through Lean

mapehe
3pts1
bytesauna.com 7mo ago

Microservices should form a polytree

mapehe
142pts119
bytesauna.com 7mo ago

Do the thinking models think?

mapehe
31pts87
bytesauna.com 8mo ago

No free lunch in vibe coding

mapehe
4pts2
bytesauna.com 8mo ago

My Favorite Math Problem

mapehe
2pts1
bytesauna.com 8mo ago

My Favorite Math Problem

mapehe
61pts52
news.cornell.edu 8mo ago

'Roadmap' shows the environmental impact of AI data center boom

mapehe
3pts0
bytesauna.com 8mo ago

LLMs are steroids for your Dunning-Kruger

mapehe
4pts1
bytesauna.com 8mo ago

I love boring business ideas

mapehe
1pts0
bytesauna.com 8mo ago

AI can code, but it can't build software

mapehe
4pts0
twitter.com 3y ago

A short film written and directed by ChatGPT

mapehe
3pts0
gist.github.com 3y ago

The type of `new Date() ${operator} 0` depends on the choice of the operator

mapehe
1pts0
www.catb.org 4y ago

Why Not C? (2003)

mapehe
2pts1
mapehe.github.io 4y ago

Show HN: Some WebGL Shader Action

mapehe
3pts1
github.com 4y ago

Show HN: An interactive WebGL shader using xemantic's shader-web-background

mapehe
1pts0
gist.github.com 4y ago

Immutable Data Without a Blockchain

mapehe
8pts3
viznut.fi 4y ago

Digital esthetics, environmental change and the subcultures of computer art

mapehe
16pts2
geekring.net 4y ago

Geekring

mapehe
2pts0
news.ycombinator.com 5y ago

Ask HN: What's a potential startup idea in your industry?

mapehe
1pts0
news.ycombinator.com 6y ago

Ask HN: Would it make sense to decentralise social media with PGP?

mapehe
2pts2
mapehe.github.io 6y ago

Show HN: Ray tracing 3D metaballs in real time

mapehe
3pts1
mapehe.github.io 8y ago

Blockchain for the non-engineer

mapehe
2pts1
mapehe.github.io 8y ago

Solving OpenAI Gym's CartPole-V0 with TensorFlow

mapehe
4pts0
www.integersnake.com 9y ago

Show HN: Integer Snake – Connect squares with the same number in them

mapehe
30pts10
www.integersnake.com 9y ago

Show HN: Integer Snake – Connect squares with the same number in them

mapehe
1pts1
globlog.xyz 9y ago

Hilbert's Hotel

mapehe
2pts0
[dead] 4 months ago

Hi, this is my blog. Hope you like this week's post.

Well, upon a closer look, one notices that the chessboard coloring is not necessary for the problem statement. It's kind of a hint actually as you could equally well just consider a blank 8x8 board and realize that this coloring arugment works. I just feel the problem is unreasonably difficult that way.

The coloring is kind of additional structure that is applied on the object you are working with. And I think this idea of "applying structure" is a very generic. You can solve similar combinatorial arrangement problems that way, but it goes beyond that.

I think that a nice, classic (and significantly more advanced) example is showing that plane and punctured plane (a plane with one missing point) are topologically different. The fundamental (homotopy) groups of these spaces are different, and hence the spaces cannot be continuously deformed to each other.

Somehow the spirit is the same, I feel. In this topology proof it's not a grid you are working with, but a topological space. And the structure you apply is not a coloring, but something quite abstract (a homotopy group). The idea in both cases is similar, though: You apply structure and this structure reveals something that's not easy to see directly.

The magic part is figuring out the structure that produces the data you need.

Hi HN,

I'm Matias. I run a small business (ByteSauna) with a blog on the site. I try my best to serve well thought out content. Here's this weeks post.

Hope you enjoy it!

I wrote essentially the same post yesterday. The interest in crypto is directly proportional to the BTC/USD rate. I don't see real innovation compared to 2016 when very few people talked about this stuff. The single thing that has changed is the price.

Either this question would have to phrased in a precise mathematical form, or its hopelessly in the realm of metaphysics, where you can't get anywhere beyond speculation. Without precise formulation, it's impossible to be sure that we are even talking about the same thing.

It's not unheard of that a seemingly metaphysical question can be made testable with some ingenuity. Local realism is a great example: Local realism used to be subject to metaphysical debate, but since the 60's the Bell inequalities demonstrate a contradiction between practically testable predictions of the quantum theory and local realism. These days empirically measured Bell inequality violations are standard stuff that appear in physics journals (see e.g. [1]).

Unless we get something we can measure, the simulation theory has no better basis than the existence of an invisible pink unicorn whose job is to produce the quantum effects by disturbing our lab equipment.

[1] https://www.nature.com/articles/s41567-019-0507-7

I think this is really difficult thing to explain to someone with zero background in technology. I did my best with this post, let me know if you have improvement suggestions or other comments. :)