It’s a well known fact common idea vague theory that science adapts itself to self-fulfilling prophecies from various sources. For example: Moore’s law was one of the bolder moves in science: from a set of very few examples, he extrapolated a theory that the number of transistors on integrated circuits would double every 18 months. This seemed to actually apply, and slowly it became a self-fulfilling prophecy; the industry adapted itself to the requirements of the law.

A more recent example seems to be Harry Potter (or rather, fairytales that might linger subconsciously in scientists’ minds): a group from Duke University have developed blueprints for an invisibility cloak. This is quite different from the retro-reflective material developed by Tokyo University a while ago; instead of projecting the background onto the material, the material itself is manipulating lightwaves. In nature, all materials have a positive refraction index which manipulates light (and other electromagnetic waves), bending it in ways we are used to. However, by creating materials with a negative refraction index, electromagnetic waves will bend at an unlikely – and unnatural – way. This is used to create metamaterials, and one such material can apparently be used to create invisibility.

Proper invisibility is pretty amazing, and I can’t help but feel that this is either:

1. A very delayed April fool’s, or
2. Way too much optimism over a theoretical possibility

I read up a tiny bit on the subject, and I found a nice tutorial about negative refraction indices. Unfortunately, metamaterials are only glanced over; I guess the subject is too deep to offer a cursory glance at. I did find an interesting paper that questions the validity of experiments made with microwaves onto metamaterials; it’s from 2002 though, so I assume that the scientists responsible for the invisibility plans have sorted out all the questions, or chose better materials for empirically testing the theories. I guess I oughtn’t be too sceptical about this after all.

Back to the topic of self-fulfilling prophecies: I noted an amusing reference to Star Trek technology in the first article. Interestingly, they claim that the things we see in Star Trek don’t violate any laws of physics, but I assume that they only refer to a few specific cases. After all, other people don’t seem to agree with the validity of all Star Trek physics. (But that’s of course an old link, so who knows what has happened in Physics since then.)

In the article they also mention other applications of metamaterials. Personally, I’m a pragmatic guy who thinks of common household applications: absolutely sound-proof walls that would do wonders for recording audio (and keep the neighbours happy), for example. I wonder if metamaterials could be constructed to handle different wavelengths differently – then there could be materials that let light through, but deflect electromagnetic radiation from monitors, or microwaves. Very handy stuff!

And it sounds a little like transparent aluminium. Star Trek IV anyone?

Veridict, a spinoff company from research at the Royal Institute of Technology in Stockholm, has apparently produced a robust speech recognition system that is capable of understanding human speech in very noisy environments. I.e., places where a person can’t hear the other person if they’re standing one meter away. The intended audience for this technology is ambulances, aeroplanes and other loud environments where a speech recognition system might prove useful.

I would imagine that the technology used revolves around filtering away everything that falls outside the normal human voice spectrum, identifying and removing stable noise patterns (hums, engines, interference, etc) and sharp dynamic sounds that can’t be human plosive sounds, and finally extrapolating between what’s left after all that. I have absolutely no idea what techniques are actually used – I’m just guessing. But if something similar is used, wouldn’t this system be excellent for normal users as well?

A workmate plays poker online regularily, and he has tested some available speech recognition systems; they are all useless. If one of his kids slams the door, the system raises $20; if the dog barks, he’s just folded when he had a straight flush. (Okay, those are completely fictive events in order to illustrate.) I have high hopes that the technology used by Veridict may be applicable for those cases as well!

But of course I can’t end on a high note since I’m a cynic. The worst problem for my workmate is the case where his wife or kids stumble in to ask him something; and no filtering can help in that case.

As a layman, I do think the problem is solvable, though.

1. The system should continuously keep track of the frequency range of the current speaker. Sure, it may vary half an octave when speaking, but if a new voice command is entered two octaves above the previous one, it’s bloody well not the same person unless he suddenly received a nutty kind of pain.
2. Use statistics. Form a Hidden Markov Model over expected commands; of course, that might not work very well. A better approach could be to form a statistical model over how the commands are entered. Check frequency, delay between words and so on, and only accept commands when they fit in within a reasonable variation range.

But then again, I know next to nothing of signal processing, so I’m sure that there are some problems with either of these methods: either the frequency varies too much, or there are no viable statistical correlations.

A friend sent me this article; I wonder if it’s a subtle hint that I ought to be more social? Anyway, the article is humorous but very interesting, so I sent it on to other people as well. There’s nothing particularly new in there, but for some it’s become a great eye opener. “It felt great to see that there are others like me,” was a comment I heard, for example.

That comment was rather surprising to me. I have taken it for granted not only that there are many introverts just like me, but also that people in general are aware of this. Maybe it’s the fact that most of my acquaintances are nerds - computer nerds or LARP (live action role-playing) nerds or music nerds or art nerds. I have no statistics to prove my theory, but I assume that nerds are more introvert than Normal People.

Another theory I have is that introverts are more creative than extroverts; at least when they don’t have to spend all their energy on being among people. Extroverts may be more perceptive and in tune with their surroundings, and that in turn may lead to great social benefits since their creative juices are applied on their surroundings. (People, social events, and so on.) Introverts can probably be very social, but I would wager that most of the time they find it easier to apply their creativity to something that doesn’t move around as haphazardly as a social gathering, or a person. Like, paper. Or a computer. Or wood.

I wonder how I would have applied my dubious creativity if computers weren’t available? Would I have become an artist? A writer? Dabbled in economy? More likely, a lazy farmhand somewhere.