Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rmrfstar
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
50 ms
·
451.
▲
by
rmrfstar
6y ago
This is more evidence that the market clearing price for labor is irrational and inefficient. Most workers are not compensated for (a) the surplus value they generate, or (b) the risk they bear in work activities. This article goes to point
452.
▲
by
rmrfstar
6y ago
This actually sounds like good actionable advice for a semi-technical person like a journalist. Still leaves you vulnerable to bios compromise (e.g. get some malware running in SMM before your kernel), but that can be addressed by soldering
453.
▲
by
rmrfstar
6y ago
I'm not a crypto nerd or an os dev. Is it ever ok to let the key leave ram unencrypted? serialize_key() seems like a recipe for doom.
454.
▲
by
rmrfstar
6y ago
Take several measurements and average them. The variance scales like 1/n_samples, which can take you from city-block resolution to building resolution quite rapidly.
455.
▲
by
rmrfstar
6y ago
> coworker saying "Oh, I bought this awesome coffee maker on Amazon last night!" This x1000. I have seen people invite others to eat lunch at restaurants that only accepted credit cards in order to elicit such a data sample.
456.
▲
by
rmrfstar
6y ago
Yes, but that does not stop the cell providers from selling your location [1]. You could also run an Android VM in the cloud and RDP to it when you want to use sketchy (edit: free) apps. This approach could have saved Bezos some trouble [2
457.
▲
by
rmrfstar
6y ago
Airnb and the cell provides are good examples [1],[2]. The cell provider location data is the most insidious. They add noise to it, but the central limit theorem is a real thing and people who buy the data are aware of that. [1] https:&#x
458.
▲
by
rmrfstar
6y ago
An RSA break only breaks key distribution. If you can securely get someone an AES key, you are all set. That is why asset managers are experimenting with quantum communications channels [1]. [1] https://www.bloomberg.com/ne
459.
▲
by
rmrfstar
6y ago
Yahoo! doesn't have data for delisted or bankrupt companies, a "survivorship bias." You need a set of companies that lacks this bias, and the set of Russell 3000 index constituents is a good set to use. Some of the older Russ
460.
▲
by
rmrfstar
6y ago
Yup. What you really want to know is "how problematic is the noise in this data"? One way to answer it is to create your own FF sorts [1] and regress your Yahoo-derived factor returns against Ken French's (which come from CRS
461.
▲
by
rmrfstar
6y ago
Care to name these "GDP draining states"?
462.
▲
by
rmrfstar
6y ago
"Need to be outlawed" and "government does not enforce" are different things. Example 1: California will not enforce non-compete covenants. What it means to enforce such covenants is that company A can have police physi
463.
▲
by
rmrfstar
6y ago
It's tough to call it self-oppression when our governance mechanisms are literally unresponsive to 90% of the public. See [1] pdf page 10. [1] https://scholar.princeton.edu/sites/default/files/mgilens
464.
▲
by
rmrfstar
6y ago
It's not always easy with MBA types. I once had a mentor with clout on a 9 figure investment committee tell me that maximum likelihood estimation is "the dumbest idea" he'd ever heard. Words like "Cramer-Rao bound&q
465.
▲
by
rmrfstar
6y ago
But how am I going to get that VC money if I don't say "deep learning"?
466.
▲
by
rmrfstar
6y ago
The whole point of the comment is that we don't know what "MK-Ultra style mind control" means because the records were destroyed. We can only make educated guesses about what is likely to have survived the program. If you th
467.
▲
by
rmrfstar
6y ago
> Yes, but this is far out of the context of this discussion Facebook's multi-class-voting share structure is an artifact of Delaware corporate law. The DGCL changes all the time . The fact that some group of investors agreed to
468.
▲
by
rmrfstar
6y ago
The book ships with Julia examples, but most of the algorithms are available in cvxpy [1], scipy.optimize [2] and matlab optimization toolboxes [3]. [1] https://www.cvxpy.org/ [2] https://docs.scipy.org/doc
469.
▲
by
rmrfstar
6y ago
This is a very important question that is explored at length in [1] and related work. [1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3128375
470.
▲
by
rmrfstar
6y ago
Governments are not obligated to lend their coercive power to enforce all agreements, particularly ones that lead to bad outcomes. See, e.g. Sherman Act. See also, bans on organ sales and indentured servitude. It is perfectly legitimate to
471.
▲
by
rmrfstar
6y ago
I wasn't suggesting that HN or its moderators kill curiosity. I was suggesting that legal secrecy and related problems do. Thanks for the thoughtful response, though.
472.
▲
by
rmrfstar
6y ago
Another amazing resource is [1]. Seriously one of the best purchases I have ever made, and I spend a lot more on books than I should. [1] https://mitpress.mit.edu/books/algorithms-optimization
473.
▲
by
rmrfstar
6y ago
> at some point admit that we want machine learning algorithms to be able to interface with symbolic rules rather than pretending that everything is a data issue That is a long way off. We find ourselves in a situation where we have m
474.
▲
by
rmrfstar
6y ago
There is some irony in claiming that the GP comment kills curiosity. Nothing kills curiosity faster or more completely than the concealment of relevant facts. Legal opinions in this case. See e.g. USSR.
475.
▲
by
rmrfstar
6y ago
This isn't necessarily an unrelated controversy. Excessive record sealing, secret legal opinions, pay-walled public records, and the fact that Federal courts are only available to large businesses while the rest of us are forced into a
476.
▲
by
rmrfstar
6y ago
It would be a crying shame to see 2600 go the way of LinuxJournal.
477.
▲
by
rmrfstar
6y ago
Tanenbaum-Torvalds [1] Node.js edition [1] https://en.wikipedia.org/wiki/Tanenbaum%E2%80%93Torvalds_deb...
478.
▲
by
rmrfstar
6y ago
We don't really know what took place in the "mind control" program, as the records were systematically destroyed [1]. One plausible interpretation is that these were mainly efforts in coercive control, rather than Hollywood
479.
▲
by
rmrfstar
6y ago
One of many examples [1]. These tactics would be familiar to the Romans. They are as old as the hills. They are apparently common enough that HBGary proposed a collaboration with Palantir to use them against Glenn Greenwald [2]. [1] https
480.
▲
by
rmrfstar
6y ago
I don't think this is a new feature? I run an out-of-band network tap and have seen only TLS 1.3 traffic for some time now.
More ›