nextos

64 ms·

901.
by nextos 3y ago
See workflows described in: https://nixos.wiki/wiki/Python .
902.
by nextos 3y ago
I have found NixOS to be a really straightforward and robust solution for Python and deep learning. However, for other languages I use, notably Julia, it is not a good choice yet.
903.
How to handle a hands-off supervisor(kidger.site)
3 points by nextos 3y ago|1 comments
904.
by nextos 3y ago
Sure, my point was that heat and battery are really good in AMD ThinkPads too. Not as good as in ARM Macs, but sufficiently good to be a compelling alternative. I have test-driven one and I could get ~10 hours of use plus it was warm and si
905.
by nextos 3y ago
Sadly Asahi, whose work is really impressive, is getting a bit stuck. Polishing the last 10% is really hard without cooperation from Apple. It would be awesome to run Linux on a MBA M2 as a first-class citizen, like it used to be possible o
906.
by nextos 3y ago
I am curious too. I have used Stan extensively. I found it extremely polished and pleasant to use. It generated very efficient samplers for particularly weird (and enormous!) hierarchical models I had. Documentation is also great. It is als
907.
by nextos 3y ago
For ethical reasons, it is not possible to perform some kinds of research with double-blind RCTs. Besides, large biobanks provide a wealth of observational data which can be explored using causal inference. One can always use a RCT on promi
908.
by nextos 3y ago
Exactly. I have reviewed a few articles for Nature Medicine , Nature Communications , and twice for Nature . In many cases, I found really absurd causal claims, plus a total ignorance of confounding and causal inference techniques. It is
909.
by nextos 3y ago
Came here to say the same thing, actually NCD can probably do much better than 78%. Li & Vitanyi's book about Kolmogorov complexity has some interesting unsupervised examples. A simple CNN as implemented in Keras tutorial can easil
910.
by nextos 3y ago
Same here. My family uses my MacBook Pro from 2009 on a daily basis, the original unibody model. Zero issues, I just had to migrate to Linux to get OS updates. It has pretty damn good reliability. Hardware can last pretty long, it is wastef
911.
by nextos 3y ago
> "The Art of Prolog" and "The Craft of Prolog" were also two of the better programming books I have read in my life. IMHO, The Art of Prolog is along with Concepts, Techniques, and Models of Computer Programming ,
912.
by nextos 3y ago
Your intuition is right. Liquid (refinement) types, design by contract and Hoare logic have, essentially, the same expressiveness. Personally I find refinement types more promising than dependent types, but time will tell.
913.
by nextos 3y ago
Also take a look at Dafny, Why3/WhyML, and Ada/SPARK. Very pragmatic tools that are getting some industry adoption. Coq and Isabelle also see some use, but they are more costly to use in my experience.
914.
by nextos 3y ago
Pluggable hybrids (PHEVs) have never been taken seriously by politicians, here in EU, and I think it is absurd. Most Europeans make short commutes, so these cars have a sufficient range for the vast majority of drives. They are cheaper, muc
915.
by nextos 3y ago
And use invariants to provide some formal guarantees. See e.g. https://dafny.org
916.
by nextos 3y ago
Some parts of PAIP might be outdated, but it still has really current material on e.g. embedding Prolog in Lisp or building a term-rewriting system. That's relevant for pursuing current neuro-symbolic research, e.g. https://
917.
by nextos 3y ago
Exactly. When I hear books such as Paradigms of AI Programming are outdated because of LLMs, I disagree. They are more current than ever, thanks to LLMs! Neural and symbolic AI will eventually merge. Symbolic models bring much needed effi
918.
by nextos 3y ago
Boyd & Vandenberghe's new book: https://web.stanford.edu/~boyd/vmls It has companion code in Julia and Python, and addresses many important applications, such as convex optimization where the authors' pre
919.
by nextos 3y ago
Probably HN is already part of The Pile [1]? I guess X is harder to scrap without permission. [1] https://pile.eleuther.ai
920.
Functional Data Structures and Algorithms: A Proof Assistant Approach(functional-algorithms-verified.org)
6 points by nextos 3y ago|0 comments
921.
by nextos 3y ago
I think HtDP is really good: https://htdp.org It's beautiful like SICP, without being too difficult or technical. It teaches one thing well, functional decomposition.
922.
by nextos 3y ago
> There are just too many ways to do things in Ruby I think this is not completely true. Ruby is conceptually very simple. It's not C++ or Perl. Everything is an object, and everything is done through method calls. There is also ext
923.
by nextos 3y ago
Monkeypatching is maybe more of a Rails thing than Ruby. I think the biggest problems with Ruby around the time Python began to take off were slowness (MRI was a disaster, slow, and full of memory leaks), plus patchy documentation (lots of
924.
by nextos 3y ago
Have you checked whether you can slow down her eye issues? For instance, in many cases of macular degeneration or for glaucoma it is totally possible to get a dramatic slowdown with an appropriate treatment. These treatments are becoming ma
925.
by nextos 3y ago
Spain has a long tradition in civil engineering and is able to reach some of the lowest costs per km in multiple infrastructures, despite a difficult landscape, and execute complex projects well. Two things really help: * Government has an
926.
by nextos 3y ago
Sure, but Ventile was designed for exactly the opposite, to prevent hypothermia on WWII pilots while they were awaiting rescue in the North Sea. It was also worn in the summit of Mt Everest by Edmund Hillary as it is a great windstopper.
927.
by nextos 3y ago
Came here to say the same thing. Paramo is great, robust, and sustainable. Another nice option, especially for streetwear, is Ventile cotton: https://en.wikipedia.org/wiki/Ventile Ventile clothes without nasty water re
928.
by nextos 3y ago
I don't think you are right. There is actually quite a lot of literature demonstrating how species present in hunter-gatherers are not present in the guts of individuals from industrialized societies. See e.g. https://doi.or
929.
by nextos 3y ago
Sure, but I don't think things are that simple. It's a huge ecosystem, once it falls into the wrong attractor it is hard to bring it back into a different healthy state. For example, some key bacteria and their symbionts may disap
930.
by nextos 3y ago
> [...] discover the factors our immune systems are missing in modern life and reintroduce them Sadly, many commensal bacteria that co-evolved with us have gone extinct due to a variety of modern interventions [1]. Re-engineering the gut

More ›