Monday, August 29, 2016

Quote of the Day

Abogin and the doctor stood face to face, and in their wrath continued flinging undeserved insults at each other. I believe that never in their lives, even in delirium, had they uttered so much that was unjust, cruel, and absurd. The egoism of the unhappy was conspicuous in both. The unhappy are egoistic, spiteful, unjust, cruel, and less capable of understanding each other than fools. Unhappiness does not bring people together but draws them apart, and even where one would fancy people should be united by the similarity of their sorrow, far more injustice and cruelty is generated than in comparatively placid surroundings.
The Schoolmaster, by Anton Chekhov 

Monday, August 22, 2016

Why Doesn't The Public Trust Science? Your Solution Is Part Of The Problem.

FLG is extremely interested in science, its definition, its methodologies, its perception among the public.   So, he obviously clicked on the link when he saw this title -- Why Doesn't The Public Trust Science?

Okay, there was a little bit of Republican bashing, but it was too bad.   There were solid suggestions, like this passage below:
Regarding scientific research, better standards of statistical significance are needed. Researchers should be required to register their research protocols in advance in virtual notebooks, to make it harder to get away with fiddling with an experiment’s design. Journals should create quotas for less interesting research, such as fact-checking other studies, which should be mandated by grant-givers.
FLG is especially keen on finding ways to fix incentives in science.   For example, finding ways to encourage replication studies and conduct basic fact checking is a fantastic suggestion.  But, unfortunately, we then get to this:

And the reason all of this is happening? Science in general has trouble communicating its findings to the wider public. Part of the reason could be that key parts of the scientific community are too old, too male, and/or just too out of touch.
 And then in the final paragraph, the final recommendation is this:

Given the concerning state of public distrust and the importance of science for the advancement of society, fixing it should be a national priority. A younger, more representative pool of scientists, combined with higher quality research, could help communicate important new findings to the world at large and restore the bond of trust between society and the scientific community.
So, after all that, the primary problem with science is that it's too old, male, and FLG'll guess probably too white?    Who the scientists are is apparently the primary problem, as quality research has been reduced to a parenthetical clause.

FLG does NOT dispute that a more diverse population of scientists would study different things and develop different hypotheses to be tested.   For example, FLG is convinced that women's health has been systemically understudied because the majority of medical researchers were male.   He's less convinced that remains the case, but it certainly was the case.

But FLG totally disagrees that the distrust of science has much, if anything, to do with the level of diversity in the population of scientists.  The entire point of science is to generate a hypothesis and then find as many ways to attempt to empirically disprove it that you can.   The scientist's gender, ethnicity, or age has nothing at all to do with what has been rigorously and empirically tested.

In fact, FLG would argue that calling for more diversity in science, without explaining how that would improve the quality of the science itself, rather than the perception of scientists, would only undermine the trust in science, as it would appear to make science a more explicitly political endeavor, rather than a knowledge generating one.

Thursday, August 18, 2016

DNC Hack

Given the horrible choices available in this election*, FLG has found the DNC hack the most interesting thing that's happened during this campaign season.  And it keeps getting interestinger and interestinger....

Here's the quick run down:

Summer 2015 - Cozy Bear, a cyber threat actor group who is almost certainly Russian intelligence, gained access to the DNC network.   They're pretty good at what they do (previous victims include the White House, State Department, and Joint Chiefs of Staff) and nobody noticed.   Seems like they did normal intel gathering stuff, minding their own business, until...

Fancy Bear, a separate cyber threat actor - most likely Russian Military intelligence, hacked the DNC in April 2016.

Neither of these is particularly surprising.   The political campaigns of the most powerful nation on Earth are a huge intelligence target.   FLG shrugs and says, "well, that's the game."   No big deal.

But then things get interesting....

Crowdstrike, a commercial cyber intelligence firm, publishes a blog post that accuses the Russians of executing the hack.  

Shortly thereafter, FLG thinks the very next day, a supposedly Romanian hacker nobody has ever heard of, Guccifer 2.0, stands up a blog and says he didn't it all by himself.  This is almost certainly a Russian intelligence PsyOps operation.   They release a few documents, but they don't really get any traction with the media.  So, they dump a bunch to Wikileaks and it starts to take off.

In addition to Wikileaks, there's also data leaking out via DCLeaks.  That site was likely stood up by the Russians.  But what FLG finds most shocking about that is the domain name was registered on 4/19/16 -- the same timeframe when Fancy Bear was hacking the DNC, which implies the Russian Military intelligence was planning to do this PsyOp campaign from the get go, not just them calling an audible because Crowdstrike outted them.

Fancy Bear also hacked the Democratic Congressional Campaign Committee.  They're leaking that stuff too, but whatever...FLG will move on to another topic that is also fascinating.

So, in the midst of all this, in addition to leaking documents, the Russians deny everything and publicly claim they've been hacked themselves.  Extensively.   Multiple departments, agencies, and offices.   They don't mention any suspects, but let's consider this a "Hey, Yankees, stop calling us out.  You do this shit too." message.

And a couple of days ago, some group nobody has ever heard of before - Shadow Brokers - aka another Russian PsyOp --- put tools and exploits from the Equation Group (the cyber actor name for the NSA) up for auction and released some, which look to FLG like NSA-level stuff.

BTW, FLG doubts the Fort Meade was hacked.  After the Snowden stuff, the Russians probably knew where to look and found a bunch of NSA hacks in their systems and worked their way back to other staging servers.   The operational security of the NSA should have been better, but given the Snowden leaks it was probably tough to keep out of sight.

Anyway, it's a new geopolitical world...

Monday, August 15, 2016

Time Horizons

Had this article bookmarked for a while...
Respondents with discount rates more than one standard deviation above the average of the sample had 29% less net wealth, a loss of around $130,000. More impatient people—similarly controlling for religion, income, race, sex, optimism and education—were more likely to smoke, drink excessively, and miss out on their flu shots and medical examinations.
The study, carried out with David Huffman from University of Pittsburgh, also found whites’ discount rates were systematically 11% lower than nonwhites’. An extra year of education inculcated a 2 percentage-point reduction in discount rates.

Time horizons!  Still want to see a research paper map discount rates to political orientation.

Saturday, August 13, 2016

Creative Commons License
This work is licensed under a Creative Commons Attribution-No Derivative Works 3.0 United States License.