HNHacker News
TopNewBestAskShowJobs

ced

2,394 karma · joined February 21, 2007

Data science freelancer in Montreal.

https://github.com/cstjean/

http://www.hackernewsers.com/users/ced.html

submissionscomments
ced··on OpenGL 3.1 on Asahi Linux
To be clear, is this a description of your experience with Asahi Linux?
ced··on Why Lisp?
Not sure about Common Lisp, but Julia (which is a Lisp in all ways except surface syntax) definitely does.
ced··on Speedcubing kid can skip last layer by learning 3915 algorithms for every case
Maybe I'll just show my algebraic ineptitude, but how do you know that g' restores "everything else"? I get that gg' = I, of course, but gT*g'? Rubik cube algebra isn't commutative, is it?
ced··on Neural Network Loss Landscapes: What do we know? (2021)
But whether all their basins are connected, or whether there are just multiple equivalent basins, is much less clear.

What do you mean by "connected"?

ced··on Atlassian is 20 years old and unprofitable
Whenever a cheaper Uber would come to my city, I'd drop them within a second.

Isn't the stickiness that drivers will only switch if customers use it, and customers will only switch if drivers use it?

ced··on An introduction to deep code-walking macros with Clojure (2013)
As an aside, Julia's MacroTools.jl defines pre- and post- code walking in four lines:

    walk(x, inner, outer) = outer(x)
    walk(x::Expr, inner, outer) = outer(Expr(x.head, map(inner, x.args)...))

    postwalk(f, x) = walk(x, x -> postwalk(f, x), f)
    prewalk(f, x)  = walk(f(x), x -> prewalk(f, x), identity)
It's dead simple and obvious in retrospect, but I don't think I would have ever thought of that! I've always wondered if there's a lineage to this set of definitions. Does anyone know?
ced··on Ten years after the Higgs, physicists face the nightmare of finding nothing else
> In 1973, professor Sir James Lighthill was asked by the UK Parliament to evaluate the state of AI research in the United Kingdom. His report, now called the Lighthill report, criticized the utter failure of AI to achieve its "grandiose objectives." He concluded that nothing being done in AI couldn't be done in other sciences. He specifically mentioned the problem of "combinatorial explosion" or "intractability", which implied that many of AI's most successful algorithms would grind to a halt on real world problems and were only suitable for solving "toy" versions.[15]

> The report led to the complete dismantling of AI research in England.[15] AI research continued in only a few universities (Edinburgh, Essex and Sussex).

https://en.wikipedia.org/wiki/AI_winter

ced··on Milan confirms new cycling network linking 80% of the city to bike paths
Were there lots of other children doing the same? Living in eastern Canada I've never really seen what you're describing.
ced··on Milan confirms new cycling network linking 80% of the city to bike paths
Are there bike paths, or does he ride on the road alongside cars?
ced··on E.O. Wilson has died
"Karl Marx was right, socialism works, it is just that he had the wrong species"

It seems that what Marx called socialism is basically communism.

By 1888, Marxists employed socialism in place of communism as the latter had come to be considered an old-fashioned synonym for socialism. It was not until after the Bolshevik Revolution that socialism was appropriated by Vladimir Lenin to mean a stage between capitalism and communism.

https://en.wikipedia.org/wiki/Socialism

ced··on Proof of stake is incapable of producing a consensus
Is there any forecast of where the equilibrium happen? Say, if you're in a 1000-people closed society where cryptocurrency is the only thing used to buy stuff, what's the economic calculation for the long-term energy consumption? Seems like there should be papers about that...
ced··on Batteries included with Emacs (2020)
emacs chords always bothered me

I got carpal tunnels a few years back from emacs, which forced me to try https://ergoemacs.github.io/ . I like it a lot, it's more sensibly designed. I've found it a pain to install though...

ced··on Searx – Privacy-respecting metasearch engine
From the link:

The CEO sold his previous company's data before founding DDG. His previous company (Names DB) was a surveillance capitalist service designed to coerce naive users to submit sensitive information about their friends.

Is that a fair statement? Can someone provide more context?

ced··on Joe Ligon: America's 'longest juvenile lifer' on 68 years in prison
He admitted to murdering them. (Which is what a guilty plea fundamentally is.)

Isn't the accused sometimes advised to plead guilty (regardless of the truth of the matter) because he has little chance of winning a plea of innocence? Or is that just in fiction?

ced··on Prions
FWIW Wikipedia says:

> Prion replication is subject to epimutation and natural selection just as for other forms of replication, and their structure varies slightly between species.[19]

[19] is https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2848070

ced··on Julia 1.6 Highlights
We did it for production code installed at client sites, and it has been very easy for us. YMMV
ced··on Where are all the containers? The global shortage explained
Anticipating major industry collapse, the Carrier Alliances showed remarkable discipline in aligning supply with demand as early as February; removing trade capacity from the seas by halting vessels and shipments. By doing so, the erosion of ocean rates was kept to a minimum. Basing their reaction on potential worst case scenarios, shipping carriers artificially created under-capacity. This made available space come at a premium and therefore significantly raised the cost of container shipping, stabilizing revenues for carriers.

If I'm reading this right, container shipping is an oligopoly similar to OPEC and they all banded together to reduce supply enough to raise prices, while in a competitive environment prices would have gone down due to reduced demand?

ced··on Data Science in Julia for Hackers
What do you mean, "map values"? There is

    [sin(x) for x in vec]
    map(sin, vec)
    sin.(vec)
which are roughly equivalent? Is that what you're referring to?
ced··on Satellite imagery shows Northern California kelp forests have collapsed
https://ourworldindata.org/co2/country/china?country=~CHN

You can add US for comparison. China's emissions have been plateauing for the past 10 years, while the US is going down. Per-capita, China is still less than half of the US, so it'll take a long time before the US catches up, if ever.

ced··on Cambridge Bitcoin Electricity Consumption Index
I 100% believe that legacy finance is a massive hindrance to the sustainability of our species,

Could you please expand on that? Why?

ced··on Gradient Descent Models Are Kernel Machines
Wouldn't that (similarity matters) be very useful information to encode in the NN though? Kinda like how you can do non-convolutional deep-learning on images, but it takes exponentially more training data than convolutional deep-learning does.
ced··on Forever chemicals are widespread in U.S. drinking water
How do you feel about drinking water that went through that?
ced··on Descript – A collaborative audio/video editor that works like a doc
Some of the HN comments here are suspicious as well.
ced··on Pixar's Render Farm
> while around 3/4 x faster than CPU rendering

My understanding is that for neural networks, the speedup is much more than 4x. Does anyone know why there's such a difference?

ced··on We should administer Covid-19 vaccines as fast as possible, not reserve doses
Not really true anymore, according to https://mackuba.eu/corona/#compare.daily?pop=1&c=ca-on,ca-qc...
ced··on Anki as Learning Superpower: Computer Science Edition
Yeah, 40 new words / day is very manageable. I used my own word lists instead of Anki, but same concept. 1000 words is enough for basic interactions (1 month) and 3000 words is generally good enough for conversation that doesn't get into the vocabulary long tail (just 3 months!) In an immersion context + some grammar textbook, it makes language learning much faster than is generally assumed.

Most people that feel they need a language class, but those are terrible for learning vocabulary. Vocab learning is best done solo.

ced··on No-till no-herbicide farming system in trial since 1981
Since everybody's been using fertilizers for a long (?) time, does that mean that farmlands aren't sold for as high a price? What happens when all the soil is eroded away?
ced··on The Singularity is Near: How Kurzweil's Predictions Are Faring (2017)
"why don't you work 65% less and chill?" most people would look at you funny. It's not a choice. I have rent, etc.

I see what you're saying, but at the same time, at what annual revenue level do humans actually start "chilling" more and work less? 500 thousand? I'd be curious to see statistics.

Where I live (Canadian city), a lot of parents are working less, through a variety of arrangements (stay-at-home mom, consulting, part-time work, parental leave). Some of them have pretty low incomes. Society doesn't make it exactly easy, but it's certainly possible, it's just not yet the mainstream option. Social safety nets obviously help.

Wikipedia has this nice chart: https://upload.wikimedia.org/wikipedia/commons/3/3b/Heures_t... . Some countries have seen impressive reductions.

Anyway, I definitely agree with that:

A solution that relies exclusively on choice fails the smell test.

Wealth inequality, etc...

ced··on The Singularity is Near: How Kurzweil's Predictions Are Faring (2017)
why the prediction failed (people decided to keep working lots and buy more stuff instead of chillin) totally unsatisfactory.

Why not?

ced··on The Book of Why: The New Science of Cause and Effect [pdf]
One can construct cases in Rubin's framework that DAG can not solve and vice-versa

Example?

Page 1 of 32Next →