I was curious about this too. But the AC below gave a nice hint, so I went looking for a better explanation. Here is the blurb from the Wiki,

So it seems the PSK can be passed around without revealing the passphrase. But if I also remember correctly, the PSK is supposed to rotate (or maybe that's WPA2).

"The article carries echoes of the "profit is evil and government is good" mantra so popular lately."

Is it that, or might it be "there are business where it's probably better not to center the whole focus on profits, which is what for-profit companies are forced to do"?

Of course, your mantra is shorter and more black&white-ish, so it must be true.

"These kinds of things are (fortunately for us) regulated."

That's why, say, an oil spill tied to the need to open a platform ASAP will never happen. Or... a nuclear facility will never suffer a melt down because management prefers to turn a deaf ear to the researchers that say more expense against tsunamis is needed.

Oh, wait!

"He thinks it's just private contractors that cut corners to save money? "

While government *may* cut corners to save money, since their percived profit not necessarily is tied to pure monetary profit, private contractors, *must* cut corners to save money, since their profit is pure monetary profit.

One could point out that there are fewer instances of white males being miscategorized.

White males are just about the easiest faces to categorize. They tend to have short hair that doesn't obscure facial features or create oddball shapes that confuse the classifiers. Their skin tone makes photographing them and finding edges extracting features easier than it is with darker skinned people. White people have a greater variety of eye colors that can be used to distinguish among them. "White guy face" is just about the optimal case for this problem. If I had to come up with a worst case that was also a photo of a fairly "common" person, I'd go with "dark skinned, brown eyed person wtih long hair and facial hair." That's a pretty clean sweep of all of the variables that make this a hard problem.

Good decisions? Sorry, but releasing poorly tested software like this was obviously a bad decision. The bad outcomes were a direct result of their poor decision making.

How good does the cutting edge of object recognition need to be before it's not "poorly tested" anymore, especially when it's for a silly photo app and not a medical or military application? I never hear this type of thing from people who have actually had to solve these types of problems. The reality is that objects are going to be confused with other objects. Lots of them, once we're talking about hundreds of millions or billions of samples. Some cases will fail with great regularity and patterns. The unfortunate fact here was that the pattern happened to coincidentally have really embarrassing cultural connotations.

This is one of the things I don't miss about working in machine vision. We'd run our algorithm over a zillion images and it would correctly handle all of them save a small handful and that small handful would be filed as bugs. OK, maybe we'll be able to handle that small handful at the expense of a smaller handful next time around. But the pass/fail criteria for the tool is in its overall results, not in the outliers.

The developers building vision algorithms don't typically create their own datasets. They purchase archives of images, and a lot of these problems stem from how many samples of each type are in those archives. The Google team likely has a giant database of human faces that it works with, and the ethnic frequencies are probably either the result of choices made by whatever origanization compiled it (and for whatever reason they compiled it) or the ethnic breakdown of the userbase of some app they used to grab the data. It's extremely unlikely that either of those will produce the same number of samples of every ethnic type.

It's also one thing if this was a program just designed to distinguish between different people. But it looks like it's trying to recognize objects of all sorts and distinguish between people and just about everything else. That's a hard problem, and the only response to this sort o thing is to take a regular failure case and feed it back into the training data so you can hit the next regular failure case. Hopefully it will be less coincidentally embarrassing, but it will definitely be there. Perhaps confusing bald men with balloons or something like that.

But I also think people underestimate how much skin color affects machine vision problems. I spent years in the biometrics industry and one consistent fact is that people with darker skin just don't provide as much easy-to-recognize detail as people with lighter skin. There will be more misclassifications as long as the image is taken using the visible spectrum. To a computer extracting features, dark skinned people and gorillas are both human-ish face shapes with a particular color range and somewhat indistinct geometry due to weak contrast and shadows. Distingushing between those two sets just isn't as easy as distinguishing between fair-skinned blondes and gorillas. You can make that decision just by looking at the color histograms and not even bothering with geometry.

Well, here's a good place to start. RC4 has a number of vulnerabilities and while each of them can be mitigated to a certain extent (changing keys, discarding the beginning of streams, etc), the confidence is low that implemented systems will successfully avoid all of them and not open up new vulnerabilities in the process.

Well, you have to factor in the Iranian cultural mania for disagreeing with each other. The Shah couldn't keep them under his thumb, neither can the mullahs, who have their hands full disagreeing with each other.

From a tyrant's perspective Iran is ungovernable, which doesn't mean elements in the government don't give tyranny a go on a regular basis. It's an ideal setup for producing martyrs. The futility of cracking down means you have a little space to rake some muck before official anger overcomes reason.

"But what I'm getting at is when the government pays employees way above the per capita income of taxpayers."

I see your point, but I think is moot both in practical and ethical grounds because tax-to-wages is not a one-to-one map.

In practical, it's just a market, offer and demand: if you want the best (not that you *must* want them, maybe your needs are not up to the best and brigthest but *if* you need them), you need to have three things:

1) Have a selection process to filter out everybody but the best.
2) Have a process in place to detect the mistakes on point 1 above and/or those that, still being the best when hired, may be not the best now.
3) Offer the highest pay (not necessarily just money, other perks included) so the best apply to your selection process to start with.

And with regards to ethics, I don't see how someone can be comfortable paying, say, a bartender the same than to a, say, brilliant doctorate in something you really feel useful. It is not as if that any single doctor needs to be payed in full by any single bartender's taxes. It is still reasonable for an hypothetical town of "humble farmers" to pay their, say, doctor, or judge, or sheriff, above their own average income if they feel their value to the community is also above their own average.