15 ms·
4. The NSA is incompetent, and correctly concluded that NK is responsible using erroneous reasoning.
by samolang 12y ago
4. The NSA is incompetent, and correctly concluded that NK is responsible using erroneous reasoning.
- famousactress 12y agoHaha! Good catch. I've seen people unwittingly leverage this basic pattern into entire executive careers :(
- mikeash 12y agoIt's basically a specialization of the base rate fallacy. If, say, idiots can still be right by chance 10% of the time, and experts are right 90% of the time, but only 10% of the population are experts, then lucky idiots will make up fully half of the population of people who are right.
- chavesn 12y ago10% is pretty generous for most of the work experts actually do, though. Experts do many things, one of which is to occasionally make predictions, where 10% might be more reasonable. But, for work like imagining how a given goal can be realized, I would guess an "idiot" could compose an achievable plan less than a fraction of a percent of the time.
- mikeash 12y agoI'm not sure. Computer experts? Yes, hard to get lucky there. Business experts? Luck seems much more significant.
- kbenson 12y agoIt's all about having enough variables in the system so you can point to something else as the cause when there is failure, and yourself as the cause when there is success. With willful ignorance of reality, idiots can have a success rate much higher than 10%. :)
- AlyssaRowan 12y ago5. This is an FBI statement. As the Sony movie studio's network is not a classified US Government network, it isn't the NSA's job to defend it. So, they may not have even been consulted on this - or (far more likely) their contributions, if relatively inconclusive, may have been cherry-picked for only those specific points which supported a politically-convenient, face-saving conclusion. (Which, as we all know, has happened before on a few infamous occasions.) When you have a big hammer, sometimes everything looks like a nail. When you have a database of every IP address which sends packets transiting a collection point to every other IP address suspected of being a C&C (and the NSA & GCHQ do have exactly that), everything looks like a potential controller. There is a remarkably strong bias towards false-positive confirmation caused by (amongst other things) P2P networks and UDP packets with forged IP addresses. A bias the NSA (and/or GCHQ) would warn about: a warning that, alas, law enforcement tend to not always take to heart - which is (anecdotally) partly why GCHQ rather dislike working with the plod (and quite probably the same feelings persist in the NSA towards the FBI). I still think the links to DPRK are very weak, if they even point that way at all - not that the North Korean government aren't vile dictators (they are) but because everything I've seen makes it look more like low-rent organised crime - indeed, I gather the payload and C&Cs used, which the FBI have (IMO) erroneously used in their attribution, are publicly available (though no, I will not link to them)! What I've seen even (albeit weakly) indicates at least two actors with different names, one of whom asked for a monetary ransom, and the latter mentioned the movie after the media did and thanked the other (for giving them access?). It may be (but there is no strong evidence either way) that the latter is actually the DPRK - in which case their "cyber-army" looks like low-rent organised criminals, which I admit would not be implausible, but speaks volumes about Sony's stunningly negligent incompetence! The one thing I feel absolutely confident saying is this: Sony Pictures were an extraordinarily soft target, and this was not a display of any high degree of technical competence on the part of the attackers. It could've been anyone from North Korea to a disgruntled ex-employee (of which they have no recent shortage) to some random 14-year-old angry kid, or anyone in between. My sodding cat could have hacked them. ¬_¬
- yohui 12y agoAh, the Gettier case: https://en.wikipedia.org/wiki/Gettier_problem https://en.wikipedia.org/wiki/Gettier_problem