I remember the day I learned to estimate.
We were in 4th grade computer lab, and the lab teacher was showing us a math game. We clicked through a couple frames and then got to a hard problem. She asked the class who would be the guinea pig. I had already made my mark as the math kid, so everyone volunteered me. The teacher clicked Next.
Within an accuracy of 5, how many fish are in this picture? You have 10 second. Go.
It was a cascade of a-HA! moments for me. I realized, first and foremost, that I didn't have time to count each fish and I needed a different strategy. I then looked at the image and noticed that the fish were distributed pretty evenly; I could mentally divide the screen into a 2x4 grid and could pretty quickly assign 5 fish to each grid. I saw that this wasn't absolutely accurate, but it seemed close enough. I typed in 40. Success! Everyone else (even the teacher) marveled.
"How did you count them so fast?"
"I didn't. I estimated."
The key takeaways positively impact me all the time. The notion of assessing what's needed and when a solution is good enough help me apply the right level of effort/thought to problems and then just move on. The actual strategy of bucketing and being comfortable working with imprecise quantities ties into this. I'm redoing the deck, how many pounds of screws do I need to buy? We're mulching the backyard, how many bags do we need? How much will it cost? I don't need a quote to the cent, just an idea if it's $50 and we should just do it now, or if it's $500 and we should think about other options.
Estimation gives the power to quickly and cheaply assess options and formulate a plan.
Sunday, August 25, 2013
Sunday, August 11, 2013
Personally identified shoppers
Recently, there was a blurb about Nordstrom following people through their stores by tracking them using NFC or wifi or whatever. I think they were trying to see what shopping patterns users had, what order they looked at things in the store, etc.
Now, suppose we tracked who each of these people were and have their shopping histories.
Let's also suppose we add some facial recognition tech so sales consultants know who they are talking to. In Marisa's case, the sales person could know to recommend new purses or jeans, and even the price range to focus on.
Seems like a win, right?
Now, what if each interaction were analyzed to see how well sales techniques, pressure (or lack thereof) translate into sales. Now, what if instead of Nordstrom doing this themselves, they bought this information from pesky magazine salesmen and so on. Now, they would know which shoppers to focus on and who they can pressure into sales.
I don't mean to pick on Nordstrom here, this could be anyone. But, this is a new arena of privacy that could be exploited to tangible, dollar benefits ... and not one that we really think of. Is my personality private? Is it something that I need to protect?
Now, suppose we tracked who each of these people were and have their shopping histories.
Let's also suppose we add some facial recognition tech so sales consultants know who they are talking to. In Marisa's case, the sales person could know to recommend new purses or jeans, and even the price range to focus on.
Seems like a win, right?
Now, what if each interaction were analyzed to see how well sales techniques, pressure (or lack thereof) translate into sales. Now, what if instead of Nordstrom doing this themselves, they bought this information from pesky magazine salesmen and so on. Now, they would know which shoppers to focus on and who they can pressure into sales.
I don't mean to pick on Nordstrom here, this could be anyone. But, this is a new arena of privacy that could be exploited to tangible, dollar benefits ... and not one that we really think of. Is my personality private? Is it something that I need to protect?
Tuesday, August 6, 2013
Joys of home ownership
I bought my Tucson home in November 2006, then promptly moved to Seattle. I rented it to a steady stream of intertwined people through May 2013 at which point they left and the house was vacant for the first time. I was busy with other stuff so I lapsed a bit on showing it and getting it ready. That's when the fun began.
1. A burglar! I got a call one morning that someone had hopped the back fence and started hacking off copper piping from the yard. Ugh. Seriously, this is like the meth-head moment of the day. The mount of copper taken might net $10-20 in scrap. Of course it caused hundreds of dollars of damage to my house (and also to the water bill). And to prove his brilliance, the guy came back a few days later to finish the job. The neighbor saw him, called the cops, and hopefully he's now in jail.
2. After getting the burglar damage fixed up, I called in a property manager to handle the house. She told me there were a large number of issues with showing the house. It wasn't really clean, there was a broken window, there were a number of other issues and so on. She wasn't willing to handle coordinating all this, so I had to ask my dad to help. Bit by bit, we got the big stuff handled, but that ate up most of July.
3. I called her back to let her know we'd gotten it in pretty good shape. She checked back, confirmed this, then told me that she couldn't represent the house because she'd noticed a bow in the floor. She hypothesized that it was a tree root pressing up against the laundry room. While in town I looked, and it was pretty obvious that there really is an issue. So, now I need to have my dad go back and babysit yet another contractor and hope there's not a huge cost associated with this. Ugh.
In short, I've spent the last 2.5 months not getting the house rented out.
Updates:
4. Finally tracked down a new property manager. I talked with them on Monday (Aug 19). They then took over the keys from the previous person, checked out the house, noted some issues and sent me a pile of paperwork to sign. This is as much as we've accomplished in 6 days. No concrete work list yet. No concrete dollar amounts yet.
1. A burglar! I got a call one morning that someone had hopped the back fence and started hacking off copper piping from the yard. Ugh. Seriously, this is like the meth-head moment of the day. The mount of copper taken might net $10-20 in scrap. Of course it caused hundreds of dollars of damage to my house (and also to the water bill). And to prove his brilliance, the guy came back a few days later to finish the job. The neighbor saw him, called the cops, and hopefully he's now in jail.
2. After getting the burglar damage fixed up, I called in a property manager to handle the house. She told me there were a large number of issues with showing the house. It wasn't really clean, there was a broken window, there were a number of other issues and so on. She wasn't willing to handle coordinating all this, so I had to ask my dad to help. Bit by bit, we got the big stuff handled, but that ate up most of July.
3. I called her back to let her know we'd gotten it in pretty good shape. She checked back, confirmed this, then told me that she couldn't represent the house because she'd noticed a bow in the floor. She hypothesized that it was a tree root pressing up against the laundry room. While in town I looked, and it was pretty obvious that there really is an issue. So, now I need to have my dad go back and babysit yet another contractor and hope there's not a huge cost associated with this. Ugh.
In short, I've spent the last 2.5 months not getting the house rented out.
Updates:
4. Finally tracked down a new property manager. I talked with them on Monday (Aug 19). They then took over the keys from the previous person, checked out the house, noted some issues and sent me a pile of paperwork to sign. This is as much as we've accomplished in 6 days. No concrete work list yet. No concrete dollar amounts yet.
Renting the Chevy Equinox
We had reserved a Chevy Impala or similar, but upon arrival to Sky Harbor were told that they were out of all full-size cars and we would be upgraded to a crossover for free. We got to know this beast over the next 5 days and 800 miles. There were no markings on it otherwise, so I'll assume it was the front-wheel-drive, 2.4L model.
Pros:
1. Lots of space for water bottles. The Equinox has lots of nooks and crannies which seem just about correctly sized for standard plastic bottles.
2. Very good A/C. This was particularly vital in August, in Arizona.
Cons:
1. Not materially larger than a VW Jetta. Sure the whole thing was up higher, but the interior height, legroom and trunk space weren't noticeably bigger that Marisa's 2007 Jetta.
2. Bizarrely deep center console. Like, down into the bowels of the car deep. Like, you can lose stuff into it deep. Why?
3. Very non-responsive gas pedal. This seems to be a hallmark of GM vehicles in particular. The transmission seems to divert all power to some mystery vortex, in effect giving the gas pedal 3 modes: off, hold speed and full throttle. This was particularly bad when driving up hill because "hold speed" actually translated into "slowly bleed speed", meaning I had to go back and forth between that setting and full throttle to keep my pace roughly constant.
4. Bizarre brake pedal placement. The pedal was at least 8 inches up from the floor of the car, meaning that if I had my heel on the floor, I could barely reach the brake pedal with my toes. Huh? Why? I literally had my foot slip under the pedal once when I was trying to brake.
5. Not very good seats. Everyone mentioned that the seats weren't exactly supportive or comfortable for longer periods of time.
6. Once I got up to highway speeds, I realized that the steering wheel is too low and obstructs the view of the top the speedometer. No idea how fast I was going unless I ducked down. I've noticed this in other GM cars (Saturn Sky, for example) as well.
7. At highway speeds, we noticed some rattling. This was not road noise, just poor construction quality in a car with 18,000 miles on it.
8. The car's suspension was also quite soft. This is expected from a soccer mom car, but the resulting bobbing and swaying did not play well with passengers prone to motion sickness.
9. Rolling the window down at highway speeds resulted in a very loud, throbbing WHOOP WHOOP WHOOP sound I'd never heard in any other car before. Unfortunately a lower volume variant of this was present even at 40-45mph, which is a totally normal speed at which to have the windows down.
10. To get the window back up (or down), I had to hold the lever the whole way.
11. When it turned to night and I flipped on the headlights, I had to find the inch-wide ring on the control stick to twist, instead of being able to grab any part of the end of the control stick and twist the whole thing. Why? The remaining, non-moving, end piece of the stick served no other purpose.
12. The dials to control things like the A/C fan had no physical settings, they were like a radio knob that lead to a digital readout (which was only shown right after a change in knob setting). Hence, we never knew what setting it was on. Not a huge deal, just kinda curious.
13. The rear visibility was poor. This is expected, and it does have a rear camera. However, that camera has no overlay so it's not really clear how wide said view is, what I may or may not hit while turning, etc.
14. Odd seat fabric. It has an overlay similar to a drawer liner, so every time you sit on it in shorts you leave with a fine fishnet pattern imprinted into your legs.
15. Easily left in the dust by a Honda Odyssey. Minivan soccer mom kicked our ass!
Rant:
While many of the above are quibbles and not really a big deal, I'd like to sound off about a few of these.
The brake pedal is flat-out unsafe. If my size 10.5 foot can occasionally slide under it, imagine what happens with your wife or daughter's size 6 feet.
The gas pedal issue is just bizarre. Every non-American company (Ford has gotten better) has figured this out. It's not a matter of how much power a car does or doesn't have, it's about actually putting it responsively to the road. Ideally (and with many other car makers' vehicles) a slight press of the gas pedal would result in a slight acceleration. Instead, the gas pedal has to be pressed pretty far down before the car actually downshifts and starts to accelerate at all. The Equinox has a 6-speed automatic transmission, which in some sense just makes the problem worse because on even a slight uphill the car has to downshift twice. The car is 3800lbs and has peak power of 182hp, so even in an ideal world it's not going to haul ass. On top of this, peak torque is at 4900rpm, which is a speed the engine really doesn't like to make it up to. This is just bad engineering. Bad GM, bad.
There's really no incentive to buy this car over, say, Marisa's Jetta. I don't think we would have been noticeably less cramped in it, would have gotten better overall gas mileage and could have actually accelerated. GM needs to wake up and realize that better engineering and driving characteristics are not something a lot of people ask for, but it's something they notice when they do drive.
Pros:
1. Lots of space for water bottles. The Equinox has lots of nooks and crannies which seem just about correctly sized for standard plastic bottles.
2. Very good A/C. This was particularly vital in August, in Arizona.
Cons:
1. Not materially larger than a VW Jetta. Sure the whole thing was up higher, but the interior height, legroom and trunk space weren't noticeably bigger that Marisa's 2007 Jetta.
2. Bizarrely deep center console. Like, down into the bowels of the car deep. Like, you can lose stuff into it deep. Why?
3. Very non-responsive gas pedal. This seems to be a hallmark of GM vehicles in particular. The transmission seems to divert all power to some mystery vortex, in effect giving the gas pedal 3 modes: off, hold speed and full throttle. This was particularly bad when driving up hill because "hold speed" actually translated into "slowly bleed speed", meaning I had to go back and forth between that setting and full throttle to keep my pace roughly constant.
4. Bizarre brake pedal placement. The pedal was at least 8 inches up from the floor of the car, meaning that if I had my heel on the floor, I could barely reach the brake pedal with my toes. Huh? Why? I literally had my foot slip under the pedal once when I was trying to brake.
5. Not very good seats. Everyone mentioned that the seats weren't exactly supportive or comfortable for longer periods of time.
6. Once I got up to highway speeds, I realized that the steering wheel is too low and obstructs the view of the top the speedometer. No idea how fast I was going unless I ducked down. I've noticed this in other GM cars (Saturn Sky, for example) as well.
7. At highway speeds, we noticed some rattling. This was not road noise, just poor construction quality in a car with 18,000 miles on it.
8. The car's suspension was also quite soft. This is expected from a soccer mom car, but the resulting bobbing and swaying did not play well with passengers prone to motion sickness.
9. Rolling the window down at highway speeds resulted in a very loud, throbbing WHOOP WHOOP WHOOP sound I'd never heard in any other car before. Unfortunately a lower volume variant of this was present even at 40-45mph, which is a totally normal speed at which to have the windows down.
10. To get the window back up (or down), I had to hold the lever the whole way.
11. When it turned to night and I flipped on the headlights, I had to find the inch-wide ring on the control stick to twist, instead of being able to grab any part of the end of the control stick and twist the whole thing. Why? The remaining, non-moving, end piece of the stick served no other purpose.
12. The dials to control things like the A/C fan had no physical settings, they were like a radio knob that lead to a digital readout (which was only shown right after a change in knob setting). Hence, we never knew what setting it was on. Not a huge deal, just kinda curious.
13. The rear visibility was poor. This is expected, and it does have a rear camera. However, that camera has no overlay so it's not really clear how wide said view is, what I may or may not hit while turning, etc.
14. Odd seat fabric. It has an overlay similar to a drawer liner, so every time you sit on it in shorts you leave with a fine fishnet pattern imprinted into your legs.
15. Easily left in the dust by a Honda Odyssey. Minivan soccer mom kicked our ass!
Rant:
While many of the above are quibbles and not really a big deal, I'd like to sound off about a few of these.
The brake pedal is flat-out unsafe. If my size 10.5 foot can occasionally slide under it, imagine what happens with your wife or daughter's size 6 feet.
The gas pedal issue is just bizarre. Every non-American company (Ford has gotten better) has figured this out. It's not a matter of how much power a car does or doesn't have, it's about actually putting it responsively to the road. Ideally (and with many other car makers' vehicles) a slight press of the gas pedal would result in a slight acceleration. Instead, the gas pedal has to be pressed pretty far down before the car actually downshifts and starts to accelerate at all. The Equinox has a 6-speed automatic transmission, which in some sense just makes the problem worse because on even a slight uphill the car has to downshift twice. The car is 3800lbs and has peak power of 182hp, so even in an ideal world it's not going to haul ass. On top of this, peak torque is at 4900rpm, which is a speed the engine really doesn't like to make it up to. This is just bad engineering. Bad GM, bad.
There's really no incentive to buy this car over, say, Marisa's Jetta. I don't think we would have been noticeably less cramped in it, would have gotten better overall gas mileage and could have actually accelerated. GM needs to wake up and realize that better engineering and driving characteristics are not something a lot of people ask for, but it's something they notice when they do drive.
Sunday, August 4, 2013
23 and Me
Marisa and I sent in our samples, but we're still waiting on the analysis. In the meantime I've been filling out various surveys from which 23 and We can do correlation studies. One of them asked me to self-identify my ethnic background. At the conclusion of the survey, they show the aggregate data from all surveys:
What's interesting here is that both Hispanic and black are about half-represented compared to the national demographics. I wonder why this is?
The knee-jerk reaction would be to say that these demographics have less money to spend on genetic tests. Or is it that they have less interest in science-y things? Ideas?
The knee-jerk reaction would be to say that these demographics have less money to spend on genetic tests. Or is it that they have less interest in science-y things? Ideas?
Hippo-critical!
Most of us have probably seen the Fox interview with Reza Aslan where the anchor repeatedly asks him why, as a Muslim, he would want to or could properly write about a character in Christian religion.
It's "fitting" that the following cases have likely never been questioned in the same way:
1. Straight congressmen deciding what rights gay Americans get
2. Men deciding how a woman should be medically treated
3. White guys talking about the background of a black president
4. .. get creative.
The very idea that someone from one background couldn't/shouldn't have interest in someone from another background, or that they could possibly credibly understand something about them is disturbing. Are we to believe that ignoring the "others" outright is ever a way forward in the world?
** - the title is taken from one of my favorite forum comment mis-spellings.
It's "fitting" that the following cases have likely never been questioned in the same way:
1. Straight congressmen deciding what rights gay Americans get
2. Men deciding how a woman should be medically treated
3. White guys talking about the background of a black president
4. .. get creative.
The very idea that someone from one background couldn't/shouldn't have interest in someone from another background, or that they could possibly credibly understand something about them is disturbing. Are we to believe that ignoring the "others" outright is ever a way forward in the world?
** - the title is taken from one of my favorite forum comment mis-spellings.
Friday, August 2, 2013
If, then, otherwise
In reading more and more about the anti-vaccine movement (and observing it on media and social media), I kept coming across description of parent after parent who said "I don't want to give my kids a vaccine that might harm them".
Suppose we allow that some vaccines are, in fact, harmful (this is generally found to be untrue, though some reactions can occur, sometimes with bad results like encephalitis, which can lead to permanent injuries). We are then faced with the following:
If I give my child a vaccine, they might be irreversibly harmed.
We've allowed the causation in this case, so let's go ahead and allow this statement. Now, what happens after you don't give them the vaccine? What's that you say, you didn't think about that one? Let's fill it in for you:
If I give my child a vaccine, they might be irreversibly harmed. Otherwise, they may contract a deadly disease.
Doesn't sound so great, does it. We can now argue that by living in the bubble of the USA they won't be exposed to a lot of the things we now vaccinate for, but given the amount of international travel people do it's a safe bet that measles, mumps, rubeola, diphtheria, tetanus, polio, etc are alive and well and have made it past our comprehensive TSA and immigrations screenings. If your child is the only one not getting vaccinated, maybe they're ok. But as a parent, do you want to 100%, for sure, feel (and actually be) responsible when they come down with one of the above? And when your neighbor also skips vaccines, when both your kids are now sick? Or when you convince lots of people to not get vaccines and an epidemic explodes?
Given the rare of births in the country and the frequency of settled suits in vaccine court (where the plaintiff doesn't even have to prove fault, it's really just a catch-all streamlined compensation system), we can conclude that there's 1 in somewhere between 10,000 and 100,000 chance your kid has an unfortunate, seriously bad reaction. That's slightly better odds than your kid being involved in a fatal car crash (and certainly way better than the odds of being injured or killed in a crash). In other words, you should stop allowing your kid in a car. Ever.
For the record: before broad vaccination programs, somewhere between 1 and 2 million people a year died in America from diseases that we now vaccinate against (that number is now under 10,000 per year if I remember right - that's right, vaccines as a whole are only something like 99% effective). That's something like 1 in 100. Put that in the otherwise clause above and see how you like it!
People in general are bad at looking at the other option; they tend to just look at choices in a vacuum. Every choice not only means you go down one road, but you simultaneously don't go down the other road. If, then, otherwise. There's always an otherwise.
Suppose we allow that some vaccines are, in fact, harmful (this is generally found to be untrue, though some reactions can occur, sometimes with bad results like encephalitis, which can lead to permanent injuries). We are then faced with the following:
If I give my child a vaccine, they might be irreversibly harmed.
We've allowed the causation in this case, so let's go ahead and allow this statement. Now, what happens after you don't give them the vaccine? What's that you say, you didn't think about that one? Let's fill it in for you:
If I give my child a vaccine, they might be irreversibly harmed. Otherwise, they may contract a deadly disease.
Doesn't sound so great, does it. We can now argue that by living in the bubble of the USA they won't be exposed to a lot of the things we now vaccinate for, but given the amount of international travel people do it's a safe bet that measles, mumps, rubeola, diphtheria, tetanus, polio, etc are alive and well and have made it past our comprehensive TSA and immigrations screenings. If your child is the only one not getting vaccinated, maybe they're ok. But as a parent, do you want to 100%, for sure, feel (and actually be) responsible when they come down with one of the above? And when your neighbor also skips vaccines, when both your kids are now sick? Or when you convince lots of people to not get vaccines and an epidemic explodes?
Given the rare of births in the country and the frequency of settled suits in vaccine court (where the plaintiff doesn't even have to prove fault, it's really just a catch-all streamlined compensation system), we can conclude that there's 1 in somewhere between 10,000 and 100,000 chance your kid has an unfortunate, seriously bad reaction. That's slightly better odds than your kid being involved in a fatal car crash (and certainly way better than the odds of being injured or killed in a crash). In other words, you should stop allowing your kid in a car. Ever.
For the record: before broad vaccination programs, somewhere between 1 and 2 million people a year died in America from diseases that we now vaccinate against (that number is now under 10,000 per year if I remember right - that's right, vaccines as a whole are only something like 99% effective). That's something like 1 in 100. Put that in the otherwise clause above and see how you like it!
People in general are bad at looking at the other option; they tend to just look at choices in a vacuum. Every choice not only means you go down one road, but you simultaneously don't go down the other road. If, then, otherwise. There's always an otherwise.
Thursday, August 1, 2013
NSA == Vaccines?
There's been a lot of hoopla around Edward Snowden's leaked slides which suggest that there's a massive NSA data gathering scheme in effect. But, what do we really know? So far we have the following primary data:
1. Edward Snowden's claim that these projects exist. He's an NSA contractor.
2. Edward Snowden's leaked slides about PRISM (there are supposedly 41 slides, and so far I think 8 have been leaked - why only 8??)
3. Edward Snowden's leaked slides about X-Keyscore
We then have independent reporting from The Guardian and The Washington Post based on this information, referencing nothing more than "top secret documents". We then have a bunch of me-too articles that basically rehash the above info and add conjecture based on quote snippets.
Let's start by examining the provided data for a moment. We have one whistleblower's claims and a subset of slides. The only slide that actually indicates the program exists is the one about signup dates of the various key players (interestingly, Skype was on a Sunday ... seems a touch odd, also why are the copies in the The Guardian and The Post different??). All other slides just have high-level descriptions of a system. These could be a report to higher ups about a working system, or a high-level spec used to get buyoff on or evaluate a project. Hell, for all we know the whole thing could be fake (also randomly curious: why are two of the leaked slides shown with a red-bordered logo, while the others are not like this - these things tend to come from a template). Why is there stuff redacted in the slides? Did Snowden do this? Who knows if these slides were produced by a conglomerate of people, but the slides show two different composition styles (granted this is very fuzzy): some are fairly polished looking, others look like a high schooler's first attempt at PowerPoint. Most importantly, if he has all the slides, why not release them all? What is he saving them for? Or what's he not wanting to share with us?
Now let's examine additional data. The biggest quote bundle is coming from Henry Clapper, who has stated that a) the reports contain numerous inaccuracies and b) that it's absolutely awful that data about programs was leaked. This could be taken in several ways:
1. It's awful that people now know about PRISM. The report was inaccurate because we actually call it BANANAS and it actually involves every company on earth.
2. It's awful that we can't trust people to keep secret things secret that they promised to keep secret. By having a lack of trust, how can we continue any clandestine programs safely? Also, the report is inaccurate because it's not about broad data collection, but rather a streamlined process by which we request targeted data about specific users (for which we have broad, bilateral confirmation). Also, it's called BANANAS.
The knee-jerk reaction to a lack of total denial is "ah, see, they are hiding stuff!" Of course they are, they're the freakin' NSA. Everything they do is secret. Hell, people who work for them aren't even supposed to tell people where they work. Also, definitive statements give information to the outside world. Less specific comments don't. The general policy here is "we don't comment on such matters", right?
There are striking similarities to the anti-vaccine movement that gained steam about 10 years ago. It all started with a "whistleblower" (Andrew Wakefield) who presented "evidence" (a very poor piece of research) that MMR vaccines and/or the thimerosal in some vaccines could be directly linked to the rise in autism. The media, in general, blew this story up. Not being able to find any other scientists to take Wakefield's side, they relied on "debates" and "perspectives" involving angry moms of autistic kids (guess which side they took). They took an immensely complicated scientific topic and tried to boil it down; they boiled too far and too fast, leaving only charred ashes in the bottom of the pot. They neglected to check Wakefield's background (attention seeker), his study's background (funded by a law firm to specifically go prove this link) and the scientific reaction to the study (total rejection - that it was published to begin with is some bizarre miracle). They also seized on the technically correct, but awkwardly worded, statements from the CDC and other scientific bodies who refused to say "there is proof of no causation or correlation", because scientists never say that. They can't, technically, be sure that there never has been, isn't, or never will be a causation in every single case, ever. The story, built on fear, emotion and knee-jerk reactions to bad information, spiraled out of control until the public had no idea what to believe anymore.
My main takeaway is that if the evidence seems sketchy or convoluted, or has gaps, or is in any other way not a definitive piece of data with clear context, we should be extra careful before we believe it and the downstream analysis of it: there's still a very real chance that, as the author of The Panic Virus wrote, "it turned out that it wasn't a house of cards, but that there were no cards at all."
1. Edward Snowden's claim that these projects exist. He's an NSA contractor.
2. Edward Snowden's leaked slides about PRISM (there are supposedly 41 slides, and so far I think 8 have been leaked - why only 8??)
3. Edward Snowden's leaked slides about X-Keyscore
We then have independent reporting from The Guardian and The Washington Post based on this information, referencing nothing more than "top secret documents". We then have a bunch of me-too articles that basically rehash the above info and add conjecture based on quote snippets.
Let's start by examining the provided data for a moment. We have one whistleblower's claims and a subset of slides. The only slide that actually indicates the program exists is the one about signup dates of the various key players (interestingly, Skype was on a Sunday ... seems a touch odd, also why are the copies in the The Guardian and The Post different??). All other slides just have high-level descriptions of a system. These could be a report to higher ups about a working system, or a high-level spec used to get buyoff on or evaluate a project. Hell, for all we know the whole thing could be fake (also randomly curious: why are two of the leaked slides shown with a red-bordered logo, while the others are not like this - these things tend to come from a template). Why is there stuff redacted in the slides? Did Snowden do this? Who knows if these slides were produced by a conglomerate of people, but the slides show two different composition styles (granted this is very fuzzy): some are fairly polished looking, others look like a high schooler's first attempt at PowerPoint. Most importantly, if he has all the slides, why not release them all? What is he saving them for? Or what's he not wanting to share with us?
Now let's examine additional data. The biggest quote bundle is coming from Henry Clapper, who has stated that a) the reports contain numerous inaccuracies and b) that it's absolutely awful that data about programs was leaked. This could be taken in several ways:
1. It's awful that people now know about PRISM. The report was inaccurate because we actually call it BANANAS and it actually involves every company on earth.
2. It's awful that we can't trust people to keep secret things secret that they promised to keep secret. By having a lack of trust, how can we continue any clandestine programs safely? Also, the report is inaccurate because it's not about broad data collection, but rather a streamlined process by which we request targeted data about specific users (for which we have broad, bilateral confirmation). Also, it's called BANANAS.
The knee-jerk reaction to a lack of total denial is "ah, see, they are hiding stuff!" Of course they are, they're the freakin' NSA. Everything they do is secret. Hell, people who work for them aren't even supposed to tell people where they work. Also, definitive statements give information to the outside world. Less specific comments don't. The general policy here is "we don't comment on such matters", right?
There are striking similarities to the anti-vaccine movement that gained steam about 10 years ago. It all started with a "whistleblower" (Andrew Wakefield) who presented "evidence" (a very poor piece of research) that MMR vaccines and/or the thimerosal in some vaccines could be directly linked to the rise in autism. The media, in general, blew this story up. Not being able to find any other scientists to take Wakefield's side, they relied on "debates" and "perspectives" involving angry moms of autistic kids (guess which side they took). They took an immensely complicated scientific topic and tried to boil it down; they boiled too far and too fast, leaving only charred ashes in the bottom of the pot. They neglected to check Wakefield's background (attention seeker), his study's background (funded by a law firm to specifically go prove this link) and the scientific reaction to the study (total rejection - that it was published to begin with is some bizarre miracle). They also seized on the technically correct, but awkwardly worded, statements from the CDC and other scientific bodies who refused to say "there is proof of no causation or correlation", because scientists never say that. They can't, technically, be sure that there never has been, isn't, or never will be a causation in every single case, ever. The story, built on fear, emotion and knee-jerk reactions to bad information, spiraled out of control until the public had no idea what to believe anymore.
My main takeaway is that if the evidence seems sketchy or convoluted, or has gaps, or is in any other way not a definitive piece of data with clear context, we should be extra careful before we believe it and the downstream analysis of it: there's still a very real chance that, as the author of The Panic Virus wrote, "it turned out that it wasn't a house of cards, but that there were no cards at all."
Subscribe to:
Posts (Atom)