How many here remember the discussions early in the previous decade about Admiral Poindexter and his Total Information Awareness concept? If one were to apply neural network techniques to VISA transactions, then a system might learn to identify subtle patterns that matched known terrorist events and might be usable to detect precursors to as yet unknown plots. The more data and the more different sources, the better the chance of training such networks to find patterns. Of course the numbers of false positives would be huge at first… and although it would go down over time, it would still remain fairly large as there is just too much noise in real world data and real terrorists would try to randomize their behavior after a few got caught.
I believe the concept is sound, the only problem is… it is utterly Orwellian. No, it is worse than Orwell imagined. It is the Holy Grail and wet dreams of the Checka, the KGB, the Stasi, the Gestapo and every other secret police system of the last century. The sad thing is that this has come to pass not in one of the many tyrannical states of the world… they are too incompentent to pull it off… but here, in our formerly free United States.
After much thought I have come to believe that Poindexter’s system was not rejected for funding and laughed out of congress as we thought at the time. That was nothing but a cover story as the whole thing slipped into the black world.
If this is TIA Black, we had better start challenging VISA, AMEX and all the others who process financial transactions. I predict that nearly every credit and debit card transaction in the US is being fed in along with the phone records and the google files and facebook pages and private email.
It is a virtual certainty. This isn’t 1984. It is much worse because as the techniques improve it becomes Skynet starring as Big Brother.
Could this be the most disgusting document ever released by a nominally democratic government agency?
http://order-order.com/2013/06/20/what-you-can-and-cant-say-on-twitter/
The process works both ways. We the people–or the computer expert among us–can do the same with all govt records and find out what the dirty bastards are up to. We can also fly drones–esp as they get smaller– and find out more and more of what is going on. The political/bureaucratic scum are in the goldfish bowl just like us.
Skynet didn’t tax us. It was more humane.
The twitter document sounds like the typical lawyer written warning label.
Snag, it all comes back to what I’ve been saying ever since this PRISM thing blew up (and for years before that, in fact): these social “networks” – which aren’t networks at all in the computer sense – are simply honey pots for spooks, prodnoses, and other assorted authoritarian bastards. If half the world’s communications are in one place, of course they’re going to spy on it, and tell you what you can and can’t say… because it’s easy for them.
I wouldn’t touch Twitter or Facebook with a ten-foot pole. They are not, as someone (possibly Eric Raymond) once said, Internet-shaped.
If they’re spying on credit cards, though, that’s a tougher one. Then again, the credit card system isn’t Internet-shaped either, because it was never meant to be. Bitcoin to the rescue, perhaps?
Skynet was better… at least it’s default Kill All Humans setting left nobody in any doubt. This version is more like the AI ending of the game deus ex. Nannynet will want to track us for our own good, and more than half the population either don’t give a fig or think it’s wonderful.
“Communications which do not fall into any of the categories above and fall to be considered separately (see below): i.e. those which may be considered grossly offensive, indecent, obscene or false.
So if I tweet that the capital of Lebanon is Damascus, the CPS will consider whether or not I should be charged with a criminal offence.
What kind of moron could draft a quasi-legal document such as this?
I emailed the Attorney General this morning – the CPS – is part of his fiefdom – asking him to resign. No substantive response yet…
I would think that there would be a big challenge here on false positives on terrorist plots. It might predict a million of the last five terrorist attacks.
Also how would it work with Rumsfelt’s unknown unknowns. If they had that ammount of computing power it may be better used to predict the next major earthquake or the winner of the 4pm at Rotal Ascot today.
Is there any libertarian reason why we do not apply the same data techniques to government oversight? They would be the one organization where I would say such methods would be warranted while at the same time it looks like they are going to be the last institution that they will actually be applied to.