If I was really organized here, I'd file these posts by topic as well to see how my take on things changes (or not) with time. Who knows; if it's relatively simple I may yet do it. In the meantime I shall muddle through with more inexpert musings on our economic system.
Here is a place that being completely unversed in "the dismal science" is a boon, as I am not tempted to try to explain the intricacies of how we do things. The trigger for this post was an account on Counterterrorism Blog of the likely cause of the 6 May 2010 Wall Street Market crash. It has been know since the word "go" that it was a computer glitch, and I've read enough sci-fi to know what happens when too much of your day-to-day life is run by computers. Specifically, it runs fine until something big goes wrong, and then all bets are off.
It is surprising the mundane nature of the errors which take things down, hence the title of this post. In the case of the latest Wall St crash, apparently the automated "buy" side got hung up, so the "sell" side went crazy. Being a computer, that means it tried it's damndest to do what it had been programmed to do even though the conditions that those parameters had been set for were not present. The effect of this was the "sell" algorithm pushing the prices lower and lower looking for a buyer.
People generally won't do that. A lack of people in the loop made it inevitable, and it happens too fast to stop once it has started. Where this treads on AotF's turf is the threat that this poses to our collective well-being. For good or ill (more "ill" recently) the stock markets are both an indicator and driver of economic health. I recently suggested that the slavish business devotion to increasing stockholder value is hamstringing economic growth and employment, and this doesn't make me change my mind. You could see what happened to stockholder value in mere minutes on 6 May 2010, and in many other panic selling situations in the past. If it's wiped out and remade that easily, it's obvious that it doesn't really exist. Why then is this a weapon to be used against us?
I will argue (right here, btw) that protecting the systems that run the exchanges from cyber attacks is important, but CRITICAL is protecting the systems that run our infrastructure. This was graphically (though implausibly) shown in Live Free or Die Hard, and there is concern that the Chinese (primarily but not exclusively) have contingency plans to do just that. The terrorist threat to major info systems is limited and local at worst, unless they have some major State support.
This opens the can of worms about how much damage your garden-variety terrorist attack (suicide or car bomb) can do. The biggest thing that a terrorist cell can put together is a truck bomb, a la Oklahoma City, 1995 . While tactically devastating, strategically it is a pinprick. Enough of them, powerful and hitting dozens of critical targets simultaneously, could have a physically (as opposed to psychological) paralyzing effect on government and society and give that strategic effect, but nothing that big and sophisticated could slip completely past intelligence and law-enforcement agencies.
The short version? No conceivable (conventional) terrorist attack, cyber or violent, could disable us provided we have any sort of decent and competent leadership. Hurt us certainly, but when deciding where to prioritize resources we must separate the real from the fanciful. We won't get into leadership here, just assume they are up to it when push comes to shove.
Computer systems, as powerful as they are today, are nowhere near as seamless as they appear in movies and TV. 24 is the baseline of preposterousness in terms of integration of computer systems; having access to every CCTV camera in a city, etc. is nowhere near realistic, and certainly not instantly on-demand as shown here and elsewhere. Likewise terrorists could not shut down the markets and utilities at will to cripple our society.
The best defence against "systemic shock" is essentially what we already have: decentralization. One Big Computer is the worst possible idea, and fortunately both human nature (territorial) and technical limitations will do most of the work of keeping the lights on. Hardening those nodes will be haphazard, but I am confident that under our current level of computer dependency any cyber attack losses will be local and/or transitory. In this case time will tell if I'm right; history isn't much help as a guide for computerized systems...
Here is a place that being completely unversed in "the dismal science" is a boon, as I am not tempted to try to explain the intricacies of how we do things. The trigger for this post was an account on Counterterrorism Blog of the likely cause of the 6 May 2010 Wall Street Market crash. It has been know since the word "go" that it was a computer glitch, and I've read enough sci-fi to know what happens when too much of your day-to-day life is run by computers. Specifically, it runs fine until something big goes wrong, and then all bets are off.
It is surprising the mundane nature of the errors which take things down, hence the title of this post. In the case of the latest Wall St crash, apparently the automated "buy" side got hung up, so the "sell" side went crazy. Being a computer, that means it tried it's damndest to do what it had been programmed to do even though the conditions that those parameters had been set for were not present. The effect of this was the "sell" algorithm pushing the prices lower and lower looking for a buyer.
People generally won't do that. A lack of people in the loop made it inevitable, and it happens too fast to stop once it has started. Where this treads on AotF's turf is the threat that this poses to our collective well-being. For good or ill (more "ill" recently) the stock markets are both an indicator and driver of economic health. I recently suggested that the slavish business devotion to increasing stockholder value is hamstringing economic growth and employment, and this doesn't make me change my mind. You could see what happened to stockholder value in mere minutes on 6 May 2010, and in many other panic selling situations in the past. If it's wiped out and remade that easily, it's obvious that it doesn't really exist. Why then is this a weapon to be used against us?
I will argue (right here, btw) that protecting the systems that run the exchanges from cyber attacks is important, but CRITICAL is protecting the systems that run our infrastructure. This was graphically (though implausibly) shown in Live Free or Die Hard, and there is concern that the Chinese (primarily but not exclusively) have contingency plans to do just that. The terrorist threat to major info systems is limited and local at worst, unless they have some major State support.
This opens the can of worms about how much damage your garden-variety terrorist attack (suicide or car bomb) can do. The biggest thing that a terrorist cell can put together is a truck bomb, a la Oklahoma City, 1995 . While tactically devastating, strategically it is a pinprick. Enough of them, powerful and hitting dozens of critical targets simultaneously, could have a physically (as opposed to psychological) paralyzing effect on government and society and give that strategic effect, but nothing that big and sophisticated could slip completely past intelligence and law-enforcement agencies.
The short version? No conceivable (conventional) terrorist attack, cyber or violent, could disable us provided we have any sort of decent and competent leadership. Hurt us certainly, but when deciding where to prioritize resources we must separate the real from the fanciful. We won't get into leadership here, just assume they are up to it when push comes to shove.
Computer systems, as powerful as they are today, are nowhere near as seamless as they appear in movies and TV. 24 is the baseline of preposterousness in terms of integration of computer systems; having access to every CCTV camera in a city, etc. is nowhere near realistic, and certainly not instantly on-demand as shown here and elsewhere. Likewise terrorists could not shut down the markets and utilities at will to cripple our society.
The best defence against "systemic shock" is essentially what we already have: decentralization. One Big Computer is the worst possible idea, and fortunately both human nature (territorial) and technical limitations will do most of the work of keeping the lights on. Hardening those nodes will be haphazard, but I am confident that under our current level of computer dependency any cyber attack losses will be local and/or transitory. In this case time will tell if I'm right; history isn't much help as a guide for computerized systems...
No comments:
Post a Comment