Thursday, November 19, 2009

Emotional Pair-a-dice

You will know by now that I have some interest in the Large Hadron Collider at CERN. You will also know that I am interested in random as something other from meaningless. And you will know that I have, therefore, an interest in the random evolutionary connections which lead to our existence being construed as something other from happenstance. My model is pretty much identical to my hope that the connections along my family's gene-line narrative were loving for the most part. The story holds together better that way. But you can make your own story out of thin air too!

Inevitably, I have been more than a little bit intrigued by well-qualified physicists suggesting that the strange accidents which have befallen the CERN supercollider might just be at that intersection between fate and our attempts to tempt it by teasing out the final particle which can only exist under conditions much closer toward the Big Bang than most of us would like to get.

They are doing the math, these mavericky physicists, and suggesting that if there is an actual way to insert a random wild-card into the planning process, then we should by all means do it. They even demonstrate why this is a way to improve the odds for getting valuable information back; regardless of the results of both the Hadron Collider successfully firing up, or it's being called off because some incredibly improbable roll of some dice said it should be!

Either way, we get more information than we would without inserting this wild card into the control mechanisms. We learn if we are tempting fate, I believe is what they're saying. And we can improve our modelling of the physical universe, even if the machine fails to fire up! Because if the impossibly improbable actually happens, and the wild-carding calls off the show, then we will have to consider that fact carefully before we spend a few more billions of dollars and kilowatts to try it all over again. We might even want to look a little harder in other directions for theoretical models of actual physics.

Of course, these guys are being ridiculed right out of the scientific community, and the prospects for actually inserting their wild card into the decision tree over at CERN are about the same as those for a bird to drop a baguette flake into the power station and bring down the entire operation. Which, um, did actually happen, by the way. But in operations this complex, if it wasn't a bird it would be something else, so that silly event proves absolutely nothing. I think that's what these scientists would also say.

So, you also know that my problem with machine intelligence is very simple, trivial and easy to understand. It comes down to random again. Consider how likely it is that an airplane builder could be convinced to include in her design some random factor, like a pair of dice being thrown automatically, which would, upon the incredibly unlikely event of say seven sevens in a row, cause the plane to self destruct.

That would be insane, and no-one would ever do it. Because we can calculate to some degree of incredible precision that after x number of flights there is near inevitability that one will crash. Perhaps, as a joke, it could be done if the probabilities were far enough out there, but even then, just like that Soviet end-game machine recently unveiled, it would seem incredibly irresponsible. I guess the reason they set it up in Russia was because they didn't trust the human side of their decision tree, so they wanted to inhibit the guys with the buttons.

Still, people fly planes knowing that there remains a certainty that accidents will happen. You have to build them to fly forever, though, and then, well kind of cross your fingers that at least the design isn't, hopefully, too fatally flawed. I, for one, am way more comfortable in a plane than driving my own car, and I'm a really good driver. But there's a lot more random out on that road.

Machines just can't do random. They don't do random. It would be incredibly irresponsible to program random into them, especially if our life depends on them.

Of course, people do random, which makes us pretty dangerous. We sometimes even take the pilot's seat after a drink, or watch our laptops when the landing strip is passing by. Among all the fly-by possibilities which our brain picks up from the world around us, most would agree that it's emotion which pikks (sic) up the ones you pay attention to.

And random would be fine among machines, so long as there were other machines to restrict the craziness. That's why they pull my Mom off to the side at the Peace Bridge when she goes into Canada. They're gathering data, by random selection, so that the humans won't be left alone to read her sweet face and get fooled. So, it might be smart to program in random just to get things going, say to call an airplane in for servicing, if never for it to self destruct.

But emotion has been paradigmatically (I use the term advisedly) banned from science, pretty much in the exact same way that random has. Whatever else is true, the "truths" which can be revealed by the scientific method exclude anything which is emotionally tainted.

To me, that's very much like the serpent which powers itself by eating its tail (I usually like to misspell tale here, but I'm giving you a break, remember?). You just simply can't get away from random. But you can say that there is a sane way to use random and an insane way.

And at is very fringiest remove from everyday life, the term emotion also has validity to very precisely the same extent that the term "particle" can have any validity at all in the "hard" science of physics. Where Schoedinger's cat is either alive or dead until you take a peek. Where a "particle" is about as hard as Joan Baez' voice.

I'm just sayin'

OK, off to roll the dice on the highways . . .



Just randomly now, the surveyors are here to line up my land. The guy looks like a hillbilly, but thinks like a physicist. He explained how he would surely know if I'd been playing with the marker posts, and why they wouldn't just simply update their last survey (saving suspicious me a ton of money), and how the conspiracy among satellites and laser distance measurements, mathematically corrected for perfect level, keep things a lot better in line than they used to be.

So, I'm kind of curious to hang around until he's finished. And then to tramp the land I'm about to leave, to find out really what I used to sort-of own. (a fiction the banking system allows me to believe)

Which gives me just another minute here to finish this silly tale. In some sense, that airplane I've been talking about depends on the goodwill of every single player along its line of assembly. And, let's hope, every single player is also subject to a lot of fail-safes, redundancies, and double-checks.

And at the end, the company builds its reputation on the safety of its planes. And still, if we're like Fox TV, we like to point out all the ways in which our interests aren't being taken into any consideration at all. And we end up preferring stupid people who look good and pretend to be like us to the smart people who we fear might be gaming us with their tricky language. Double *sigh*.

Going with your gut is great for deciding when to hit the road, and who to love. Not so hot for making really important decisions, right? But at the fringes of everything we do, it's still an emotional roll of the dice.

And, well, I'm just hoping we never depend on a machine, or machine-like thinking, to do our decidering for us.

Will machines ever think?

I'm hoping there will be a little bit of caring in the decision too, and not too much passing it off to fictional Jesus. Nor, for that matter to some roll of dice that technology will come around and do our thinking for us. Or make it just that easy to act like greedy selfish tycoons, each and every one of us.

1 comment:

Anonymous said...

Beneficial info and excellent design you got here! I want to thank you for sharing your ideas and putting the time into the stuff you publish! Great work!