agravier

12 ms·

31.
by agravier 3y ago
Could we fix the typo in the title? It almost phisically hurts to read.
32.
by agravier 3y ago
Ooooh and here I was, an ignoramus thinking OpenAI was mostly trying to find a way to milk the generative models cash cow. Silly, cynical me, I should have known better.
33.
by agravier 3y ago
cowsay "Reaaaally?"
34.
by agravier 3y ago
I invoke Poe's law with this one.
35.
by agravier 3y ago
One quickly gets that x-ray vision when interacting enough with chatgpt
36.
by agravier 4y ago
He is also well-known in neurology for his discovery of a particular type of cerebellar neurons, and his name is also misspelled in this field: https://en.wikipedia.org/wiki/Purkinje_cell
37.
by agravier 4y ago
Seems like a dry market would also do in this case.
38.
by agravier 4y ago
Is there a count somewhere?
39.
by agravier 4y ago
I've been using https://lunar.fyi/ to go below the min brightness and also beyond the max brightness on an M1, works very well.
40.
by agravier 4y ago
It matters to understand how things work in order to understand their behaviour and react properly. I've seen people draw conclusions from applying Theory of Mind tests to LLMs ( https://arxiv.org/abs/2302.02083 ).
41.
by agravier 4y ago
Contabo user here, no complain in a couple of years using between one and two VPS. 0 downtime.
42.
by agravier 4y ago
Ad hominem arguments don't work very well on this crowd, fortunately. Try criticizing the substance of the article.
43.
by agravier 4y ago
The idea seems cool but I don't understand how this is meant to be used in practice, for a beginner. As support for a structured course?
44.
by agravier 4y ago
What do you mean by "it crashed". No output?
45.
by agravier 4y ago
Stateless. The whole conversation is replayed in the input.
46.
by agravier 4y ago
Also in all sorts of data engineering, productionized data science, etc.
47.
by agravier 4y ago
I get that on Ubuntu after a few days of uptime.
48.
by agravier 4y ago
My take is that contrary to Curtis, Moxie is exaggerating or highlighting some extreme instances of "the worst" to make his point. I doubt he actually systematically goes out of his way to buy the worst stuff, as his point is to n
49.
by agravier 4y ago
OCaml is not slow in my experience, but we should probably look at what big teams using OCaml in a performance sensitive setting have to say. https://www.janestreet.com/tech-talks/safe-at-any-speed/
50.
by agravier 4y ago
My kids have been using it on a regular basis for a couple of years now, the core engine is stable and I didn't find any bug, and there are nice games and non-buggy mods.
51.
by agravier 5y ago
Computers didn't exist!
52.
by agravier 5y ago
I think that many yearn for an idealized life in a small, relatively isolated community, low tech and free of fiat currency. They try to imagine workarounds for the fact that it's not actually currently possible in reality. UBI, anarch
53.
by agravier 5y ago
Unless the author got the green light from Zach, I don't think it's fine that they reproduced an SMBC comic in there. Without any credit, to make it worse.
54.
by agravier 5y ago
The previous year's winners all look interesting but it would be even more amazing if there were some explanations about what I'm looking at, for each photo. Is there some page with explanations?
55.
by agravier 5y ago
Hello, I have experienced something similar recently with another tech company, where the hiring manager, HR and myself had agreed on everything including me applying for a visa au my own cost (and naturally stopping the interview process w
56.
by agravier 5y ago
It's an old game. I tried it 20 years ago or so. I'm French, so it was not totally unknown outside of Poland.
57.
by agravier 5y ago
Why are not all animals recognised as such under that law?
58.
by agravier 5y ago
I would, I find that there are lots of data transformations and non-deep modelling happening in python still. E.g string processing, Json munching, business rules if-this-then-remove, etc.
59.
by agravier 5y ago
Numerical computation libraries: numpy, scipy, ML libraries like sklearn, pytorch, and bindings for tensorflow, xgboost and many more must all work without any hiccup before data science and AI teams will consider switching away from CPytho
60.
by agravier 5y ago
I did a bit of work on the topic (which remains preliminary), trying to reinvent the wheel a little bit and also implementing a naive (but correct) LWW-graph structure, and I doing some light thinking around applications [1]. There is one p

More ›