Friday, July 5, 2013

The real revolutions

I was prompted by an article on Jaron Lanier's book, Who Owns the Future, to think about IT and its impact on society and the world. I admit to being underwhelmed by Lanier's hypothesis, via Timberg, that the future landscape of employment (or unemployment) will be drastically altered by the Internet. Its a bit like saying the sun will rise tomorrow - not particularly insightful and certainly anything but "visionary".

Lanier might be an IT boffin but he's possibly woefully ignorant of history. What he ought to be noting is how pathetically fad driven modern IT really is and how glacial its progress has been. What he really should be talking about is the real revolutions in history that have been transformational along so many dimensions, but are invisible today amongst the hype upon überhype from the IT industry.
Let's take one development as an example. Touch technology is so smart that a whole new technology - text prediction - had to be co-developed even to get it off the ground. Predictive text technology has been around for decades, but it has been hampered in being marketed by poor processing power. Its moment has come just in time to prevent a billion people ceremoniously smashing their 'smart' phones.
Touch screens are so entirely unintuitive because they require us to move our bodies in ways never intended in order to achieve a relatively simple outcome - text entry. Not only is input serial, it requires the use of a digit evolved to work opposing the thumb. It is almost entirely useless at fine manipulation. Too bad that even the average two finger typist like me can spew text onto a screen at about 5 times the pace of the average text entry by the most fluent in the art. Let's not even mention RSI.

The bottom line is that it's really stupid technology, saved by text prediction technology. A revolution it isn't; a revolution it will never be. It too will pass, like some ugly edition of Windows 2000, into historical oblivion.

It follows one of the most dramatic failures of human machine interactions - WIMPS. Windows, Icons, Mouse, Pointer. I have watched as a trench was dug on my place by a guy manipulating 3 levers on a back hoe. Even at a distance of 2 metres, he could carved the earth with millimetre precision. I have witnessed an interesting son of a Sri Lankan immigrant on a farm in western Queensland drop a kangaroo at full hop, while riding a motorbike at 25km over a dusty track, with a revolver, at 20 metres. I have heard Bobby McFerran mould his voice to sound like a violin; and watched in awe as pianists manipulated 10 digits with frightening speed to execute unbelievable manual gymnastics.

The fact is, with practice, humans are capable of genuinely breathtaking feats of skill. Inevitably it is the eye and the hand that conspire in this magic, tuned by a million years of genetic selection. Why did we ever imagine that a single digit was going to match a highly skilled set of fingers? What profound blindness led us to think that there was a productive edge to this?
The truth is, there isn't. IT is so pathetically underdeveloped that we should wonder why we even nod in recognition. Why would anyone, ever, repeat the "IT revolution" mantra?

Digital communication began about 4 million years ago when hominids walked upright and released their fore-limbs from transport duty, making them available, relatively unimportantly, to manipulate tools and most importantly to communicate. Gesture and body language was so refined that it persisted across species. Where facial movement and eye contact work with intimate communication, gesture has an advantage of being unambiguous over some distance. It is fast, clear, reflexive and effortless.
That gesture-mediated user interfaces were not the first development of IT is a travesty of history. The spectacular dead ends which humans have pursued are witness to our incredible resource richness in that we can waste valuable energy of worthless  or futile pursuits.
Digital communications outside of gesture began in 1792 with semaphore, not the first codification of text into a transmission mechanism, along with mark-up language and transmission protocols, but one that had international recognition and implementation and a degree of standardisation. It was rapidly followed by electrically mediated telegraphy in 1837 and was already widely used by the beginning of the Civil War in America. In a single generation, the efficiency of information transmission of text had increased 1000 fold. So successful was this system that it persisted for a century, changing only in the media of transmission - to radio.
Telephony, often heralded as a step forward, actually created more problems that it solved. It can be celebrated as our first massive regressive step. First, standardised communications protocols were almost entirely abandoned in favour of the confusion of conversation. Not only did telephony lack the support of paralinguistic elements critical to proper conversation, it also introduced institutionalised rudeness, where your attention to one person could be interrupted by another, regardless of the triviality of the business of the interrupter.
Furthermore, conversation relies entirely on spoken language which is, by its very nature, almost entirely undisciplined. After centuries of making writing less and less vulnerable to confusion by standardising spelling, syntax and register, telephony chose the least means of passing on information. Like its descendant predictive text, analogue communications technologies had to be greatly refined to faithfully carry a message such that a comment on your Auntie May's garden wasn't mistaken for "Aren't the Melroses gorging themselves".  
Moreover, telephony reduced the effective distance of transmission by a factor of 100 because entirely unimportant elements of speech, akin to frilly fonts on websites, had to be accommodated, resulting in the integrity of digital communication being entirely compromised.  In the competition of communication media, the faddish telephony overcame the practical telegraphy and progress was stymied for a century. Inanity became institutionalised. No matter how entirely disinterested you were in someone's life, it became rude not to ring them. Not calling a girl became equivalent to saying she was ugly.
The supposed revolution given to us by Bernes-Lee, heralded as some kind of hero, was likewise a giant step backwards. In HTML, entirely fatuous formatting trivia was encoded around real information and in HTTP, an entirely clunky protocol of data transfer was overlaid onto packet data transfer protocols that were already established and highly efficient. Hypertext has since proved steadfastly unintuitive since it almost entirely ignores the underlying structure of information, especially visual information.
Fortunately, back-end SQL (Structured Query Language) has saved both HTML and HTTP from themselves, as has XML, developed under pressure from frustrated information managers who found HTML to be just so much frippery. We never needed HTML when we already had portable document formats and we have yet to see a proper information linking technology.
Why isn't every link on the web fundamentally tied to every other piece of information that is semantically equivalent by the structure of the mark-up language? What possible use is a mark-up language that has to advertise that it is making a link, then define the link as data and then provide the data? Why do we not have a universal tag for "PersonName", the most universally applied information? The 'WWW revolution' was simply a massive fashion show.
We can add to this litany of disasters a whole raft of technologies that have been dramatically regressive. We could talk to people on the moon with wireless in the 1960, but chose telephony, possibly the most crippled technology of the time, on which to build the WWW. Almost half a century later, we re-discovered wireless, but only after telcos had fleeced us of a massive fortune to try and make a copper network viable.
So what technologies have really been revolutionary?
Consider mining in the 16th century. A good miner could recover about 10 kilograms of ore from a rocky seam in a day. This might render as little as 10 grams and rarely more than 1 kilogram of useful metal, but only after extensive processing that was energy intensive. The use of explosives in mining, which changed mining within a single decade, increased the output by a factor of up to 1 million but generally as high as 10000 times the output.
Hidden in that statistic is a significant advancement that was revolutionary. Explosives were so successful in uncovering ore that mining could be restricted to good weather. Before the 1600s, mining would carry on through winter and murky rainy weather, often resulting in the death of miners by gradual exposure.
In contrast to the WWW, the printing press was revolutionary. Not only did it increase the speed of 'printing' by a million fold, it forced communication to adopt highly regulated forms and structures which are still vital in the 'lol' generation world. It released the pent up social disquiet of the middle class and allowed universal access to information (attributed, falsely, to the WWW).
Revolutionary, also, was the recognition by physicists that atoms were only notionally equivalent to mini solar systems and that electrons 'grouped' themselves into quanta and where located probabilistically, rather than geographically. This led to so many radical reconsiderations of how matter and energy were related and 'operated' and allowed the development of laser technologies, without which most of the IT developments would simply not have happened, regardless of clever programming.
Other revolutions, such as vaccination and use of penicillin are largely forgotten, even denied.
But these kinds of revolution are rare, because most developments in technology are incremental and many do not progress the mission that they serve. Which brings me to consider the next incremental step.
A million years after gesture became commonplace, user machine interfaces are set to become gesture mediated. Not only is gesture easy to map and generally unambiguous, it frees technology of the requirement to touch or tap or wave a stupid little piece of plastic to move a pointer.
This heralds a swath of technologies. Foldable screens will not require touch capability and will develop rapidly. Virtual keyboards will hover in space in front of you and allow a return, just in the nick of time, to high speed, almost parallel, data input. 'Mouse' motions will return to being intuitive, rather than locked into 2 dimensions.
Windows will be meaningless, because the scope of attention will depend on the angle the lens of the camera is capable of, not the resolution and size of a screen. Eye movement control will die a natural death because eye movement is so immensely difficult to control (or even be conscious of) and devices will be free to interpret your movements even without any invitation (such as starting up when you sit down to work or switching off when you are clearly engaging in something else).

We are well overdue for a revolution. This one is in rejecting the idiocy of our immediate past and embracing a world which evolution gave us a couple of species ago.