A tongue-in-cheek, yet still stimulating thoughtpiece by our Afghan-Icelandic-Irish friend …

… Smári McCarthy … :

“[Disclaimer: This is a bit of a joke, written last night as I was falling asleep.]

I just arrived in London after another one of those mind-numbing long haul flights, this time from Mumbai. And in my eight hours of pneumonia induced pain I managed to watch a delightful array of films that I hadn’t gotten around to, including the fourth ‘Terminator’ movie.

Two-or-so years ago, just before it became public knowledge that this film was being developed, I was visiting MIT’s Center for Bits and Atoms when the makers of the film contacted CBA looking for technological goodies that could make the film more interesting. I’m sad to say that none of the more profound ideas tossed at them made the script, but either way, I think that the entire discussion contained an important implicit subtext which was lost on the kind of people who think that hurdling Christian Bale between flying machines whilst explosions happen is a good idea.

So I present here a short analysis of where the Terminator movies go wrong:

The first Terminator movie didn’t stretch things very much. It was a simple time travel scenario with man versus machine, a kind of crypto-luddite cumfest. It wasn’t until Judgement Day came around that the industrial model started to warrant scrutiny.

In Judgement Day we are treated to a view of Los Angeles being vaporized by a nuclear explosion. For the machines, this tactic makes sense. Take out major human outposts to diminish their numbers significantly straight off. Humans have very low tolerances for nuclear hijinx such as radioactivity, but machines, being simpler and more discreteized, can presumably take much higher doses before problems start to occur. Expose a titanium alloy to a source of beta radiation for long enough and sure enough it will melt or otherwise morph, but long before humans melt from that kind of radiation atoms in their DNA start picking up extra core elements, altering their nucleic structure, and causing their host to die a very brutal death.

This illustrates a model. Consider that for anything that is “required” for sustenance, or “must not be” for survival, there exists a continuum, and each individual occupies an interval on that continuum. The length of this interval is often called “slack”. More slack equals more likely to survive a lack of something crucial or an excess of something lethal.

Simply by comparing the average slack values and their 95% intervals for each individual species you can pretty easily discern the smartest set of tactics that can be employed by each side. The robots can go ahead and use nuclear instability, thermal radiation (metal objects tolerate high heat while humans like myself start to go all wiggly and faint when it’s higher than 45°C out), extreme climates, darkness, and that kind of thing to their advantage.

The humans on the other hand have a much better ways of dealing with machines at their disposal.

In Terminator 4 a huge 7-or-so-story evil robot thing came out of nowhere in one scene and started scooping up people. It later became a part of some sort of super-carrier aircraft. Each of these things must require a large amount of metal to build, not to mention rare earth metals, plastics, semiconductors, etc. In T-2 Schwarzenegger claims that he has a “metal” endoskeleton, without being specific as to which metals exactly. From what I’ve seen of the Terminator’s Moh’s hardness, it is most certainly an alloy of something. Either way, Ferrum is for this kind of purposes a pretty aweful atom, and it kind of only makes up for it by fact of its general ubiquity. It requires lots of special treatment to be very hard, it rusts easily, and it is a crappy conductor compared to lots of other metals.

For proper construction of a Terminator you’d presumably need a bunch of metals: Titanium, cobalt, paladium, chrome, copper, gold, silver, tantalum, etc. Each of these metals is relatively easy to get, provided you know where to look. Tantalum is a pretty good one. Most of it is mined in the Congo, by children. I would be very happy to replace those children with robots, but let’s face it: if the robots are out to kill us, one of our best ways to kill them off is to keep them away from tantalum. Even if that means making a bunch of child slave laborers unemployed. Not being able to use tantalum for capacitors would mean they’d need to use other types of capacitors, such as electrolytic, which have worse properties for a number of things, and are generally larger and more fragile.

See where I’m going with this?

Humans are part of an eco system that has been around for millenia, and through our evolution we have managed to adapt our “slack” values to be narrow for things very abundant in our environment (such as amino acids) and wide for things that are relatively scarce (such as certain metals). We can survive without tantalum. The robots cannot. We can survive without electricity. The robots cannot. We can survive without most of the infrastructure we take for granted – it won’t be pretty, but honestly, you can stick a human in a Mumbai slum far more readily than you can stick a Terminator.

Humans are good at surviving the kind of situation where everything is messed up and ugly. Our bodies adapt. Robot’s specifications don’t change. Sure, you’ll have a T-1000 liquid metal thing every now and then that’ll cause you some grief, but honestly there’s no threat that the T-1000 can pose that a little electromagnetic resonance burst can’t fix.

When it comes down to it, the battle between humans and robots is not so much about sheer power as it is about controlling the industrial chains. Attacking the slack. And as long as robots require things that are harder to get than the things humans need, the humans will win.”