md_

160 ms·

601.
by md_ 6y ago
Very cool. This reminds me a bit of Trevor Paglan's work, and especially https://www.vice.com/en_us/article/qkv8ww/the-geography-of-s... and https://www.amazon.com/Blank-Spots-Map-Geograp
602.
by md_ 6y ago
A bit more info on the legal basis for arrests (and how this does not meet that standard): https://www.lawfareblog.com/unpacking-dhss-troubling-explana...
603.
by md_ 6y ago
First, I'm aware of cases where protest organizers have asked protestors to wear black out of solidarity with BLM, so I'm not convinced by the argument that this is the entire point of wearing black. But even if we take that as a
604.
by md_ 6y ago
Certainly anyone violently attacking law enforcement is not a peaceful protestor. But protestors are not a monolithic entity: the vast majority are clearly peaceful. In a free democracy that values the right to peaceful protest, law enforce
605.
by md_ 6y ago
From the Wikipedia article: "There are also cases where the principle appears to have an effect, but weakly...In the United Kingdom, the SDP–Liberal Alliance, and later Liberal Democrats, between the February 1974 and 2015 elections ob
606.
by md_ 6y ago
I wonder if you read the linked Wikipedia entry, which includes: > In political science, Duverger's law holds that plurality-rule elections (such as first past the post) structured within single-member districts tend to favor
607.
by md_ 6y ago
Every Canadian government but for one has been from one of the two main parties.
608.
by md_ 6y ago
Professionals can be surprisingly sloppy. I mean these guys are pros, right? https://en.wikipedia.org/wiki/Poisoning_of_Sergei_and_Yulia_... And I always return to this story: https://en.wikipedia.org/w
609.
by md_ 6y ago
French Presidential elections use run-off voting; they don't use first-past-the-post.
610.
by md_ 6y ago
Hey Chris, Maybe tone doesn't come across well in written communication, but dismissing this as "clearly nonsense" without bothering to engage with the merit of the claim or, seemingly, even read the link seems disrespectful,
611.
by md_ 6y ago
The primary attribute of American politics that encourages a two party system is what’s known as “first past the post” elections. America’s two main parties have changed over the years (remember the Whigs?), so the existence of a two-party
612.
by md_ 6y ago
This is an underrated comment. What's intriguing to me is the idea that not only might the killer have been (consciously or unconsciously) role-playing a "professional", but that the police were engaging in the same fantasy.
613.
by md_ 6y ago
Agreed. But I read the parent’s comment as being about the legitimacy of such a claim itself—that the claim of political bias is not merely lacking proof but “whataboutism” and thus illegitimate or irrelevant, even if proven. :) Of course o
614.
by md_ 6y ago
Sure, but the claim here wasn't (as I read it) that this is unusual, but that it's unjust. If prosecutors are selecting the highest ROI targets based on some calculation of which cases they can win and which cases have the biggest
615.
by md_ 6y ago
I’m not sure that’s inherently disqualifying. When it comes to questions of law and justice, it seems entirely reasonable to ask about similar cases, simply to investigate the equal and fair application of law, no? “Whatsboutism”, as I
616.
by md_ 7y ago
Why do youth the variance in returns had increased? Because of the first point? I don’t think I buy the first point, though. Intellectual property rights are probably at an all time high. Is the point that many types of innovation are outsi
617.
by md_ 7y ago
That's an argument. But it seems to me still a far cry from arguing that cheap capital leads to less innovation--in effect, that dumb money drives out the smart.
618.
by md_ 7y ago
"Money (in the form of buying power) can't be created from nothing..." But you weren't talking about money creation. You were talking about interest rates. Is it your point that capital has the same cost all the time, an
619.
by md_ 7y ago
That's an interesting point. Yes, for economic growth to be measurable, the nominal gains (i.e. dollar-value gains) have to be exponential--but there are a few reasons to think that as the economy grows, innovations that lead to produc
620.
by md_ 7y ago
Forgive me, but I don't understand your point. Is it that it's so easy to borrow money that there's no need to sift good ideas from bad? And if so, that seems like it would result in more bad ideas being funded, not fewer goo
621.
by md_ 7y ago
Sorry, but where do you see breadlines?
622.
by md_ 7y ago
Yeah. By this rationale, any recession or depression is V-shaped. Which is true (or has been so far), but it doesn't help those who suffer from it. I'm also struck by the analogy to the Smoot-Hawley act and the theory that the bre
623.
by md_ 7y ago
To be slightly snarky, Rob is arguing two points here: * The WSJ allegations--that Huawei has allowed unauthorized use of its lawful intercept capabilities by Chinese intelligence--are anonymous and therefore unreliable. * Rob personally kn
624.
by md_ 7y ago
The famous experiments from Benjamin Libet seem relevant here: https://en.wikipedia.org/wiki/Benjamin_Libet#Implications_of... . In particular, see Daniel Dennett's response (the relevant aspects of which are quote
625.
by md_ 7y ago
Well, again, if that’s total annual comp, sure, for the reason I gave (vesting of previous grants biasing toward longer tenure at level, signing cliffs). But as a one-time manager at a FAANG, in my experience, the ranges for comp in terms o
626.
by md_ 7y ago
I’m a little surprised at that. FAANGs tend to have fairly prescribed comp systems. There’s flexibility around equity and signing bonuses, but if you told me that you were making (base+bonus) say $250k and someone at your level was making $
627.
by md_ 7y ago
Standard primitives, perhaps, but has the scheme itself been reviewed by competent cryptographers? At best, the Telegram developers are well meaning but have demonstrated in previous versions of MTProto that they lack a background in crypto
628.
by md_ 7y ago
As the author noted after you posted this, it’s not a given that the ML algorithm is O(n). It may be constant time (by looking only at column headers and a sample of data, say). That said, I was really more interested in practical runtimes.
629.
by md_ 7y ago
How does that compare to just doing compression on a sample of data? :)
630.
by md_ 7y ago
When and why would you want to apply ML to this problem? Off the cuff, this seems like it competes with an alternative of simply running every considered compression algorithm and choosing the optimal one. I guess this would be advantageous

More ›