Wednesday, May 30, 2012

Looking Back: Ray Kurzweil’s 2009

Originally Published In THE FIX, January 1, 2009

Ray Kurzweil’s 1999 classic, The Age of Spiritual Machines, devoted a full chapter to his predictions for the year 2009. Given that this year is already upon us, it only seems appropriate to take a look back at the predictions made for it by a thinker whose influence on science fiction has been so vast.

I will not attempt here to discuss with every one of his prognostications, some of which were vague, did not lend themselves to easy verification, or were simply commonplaces that had little to do with the book's area of concentration. Instead I will focus on the relatively unambiguous claims, particularly those having to do with specific developments he assumes to have materialized not as clunky, cranky, buggy, unmarketable prototypes forever "in development," or novelties or luxuries for people with too much money and too little sense, but to have become refined enough and affordable enough to proliferate widely. (Readers of that book will remember that there are many of these, several of which he reasserted in his discussion of the "2010 scenario" in his more recent The Singularity is Near, which appeared in 2005.) Additionally, given that the changes in information and communications technology are the linchpin of everything else here, it would only be fair to focus on those.

It must be noted that some of his anticipations proved at least partially correct, such as his observation that

• Local area networks and wireless technology will become widespread.

• The downloading of content from the web would increasingly replace the transfer of physical objects.

• There would be increasing convergence between various media via the web.

• The transition from 2-D to 3-D chip technology would be clearly underway.

• A $1,000 PC ($1300 in 2008 dollars) would be capable of 1 trillion calculations per second.

However, this was more than offset by the things he clearly got wrong.

• There would be a significant reduction in the size and weight of personal computers below that of the notebooks of 1999 - a banality, except that significant in this case means that PCs could now be embedded in clothes and jewelry.

• Rotating memories (like hard drives, CD-ROMs and DVDs) would be on their way out, in favor of "purely electronic" memory.

• Speech-recognition software will make text primarily voice-created, so that keyboards will be increasingly vestigial.

• Personal computers will be equipped with facial recognition hardware and software adequate to let them recognize their owners on sight.

• Print-to-speech and speech-to-text readers will become so capable, compact and affordable as to be standard equipment taken everywhere (and indeed, eliminate many of the handicaps of blindness and deafness in the process).

• Navigational devices will enable the blind to make their way using comparable devices, with which they can communicate verbally.

• Paraplegics will walk using computer-controlled orthotic devices.

• Long-distance driving will be automated by "intelligent roads."

• Translation software (letting you speak in English and be heard in Japanese, and vice-versa) will be standard in phones.

• "Smart paper"-quality visual displays will be standard in our electronic devices.

• Eyeglasses-based displays will routinely be used.

• Most reading (as in school) will be done off electronic displays, with e-book reader-style devices becoming the most common way of reading print media.

• Speakers will be replaced by miniaturized, chip-based devices generating 3-D sound.

• Not only will "responsive and precise" Language User Interfaces be commonly available, but "intelligent assistants" with "natural-language understanding, problem solving, and animated personalities [will] routinely assist with finding information, answering questions, and conducting transactions."

• Virtual reality, while still primitive, would be a standard consumer item (replacing, for instance, 1990s-style chat rooms), lacking only "surround" tactile environments.

• Autonomous nano-machines will have been demonstrated (though they will not yet be commercially useful).

Inevitably, his guesses about the direct consequences of technological change were off - for instance, his expectation that computing technology would be so powerful and widely available that teachers effectively become "human resources" workers concentrating on the motivation, psychological well-being and socialization of their students rather than the imparting of knowledge or intellectual training as such.

Naturally, this carried over to larger, more significant mistakes at the big-picture level. Notably Kurzweil argued that, at least in the U.S., the "ten years leading up to 2009 [will] have seen continuous economic expansion and prosperity," including a booming stock market and price deflation, with the continuing tech boom (despite occasional "corrections") leading the way. He also argued that the underclass will be stable in size, and politically neutralized by public assistance and the broad general affluence.

This prediction proved to be such utter nonsense that I will not bother to tear it apart here, virtually every word I have ever published about economics attesting to how differently things have turned out. The same goes for his guesses about the future of warfare, when he said that computer and communications security would be the main mission of the U.S. Defense Department; and that when there is violence of a more conventional kind,
Humans are generally far removed from the scene of battle. Warfare is dominated by unmanned intelligent airborne devices. Many of the flying weapons are the size of birds, or smaller.
For the moment, never mind the implicit dehumanization of the enemy, or the civilians they will not easily be extricated from, and consider the sense in which he meant this: that U.S. troops will not be there at the scene of battle. Given the attention lavished on aircraft like the Predator, it may seem that he had something there - but one need only look at the massive and costly U.S. commitment of ground troops to Iraq to see that this claim is total nonsense.

Admittedly, Kurzweil's errors will not surprise those who revel in such errors for their own sake. However, we simply cannot get away from the problem of predicting the future, as pressing issues like climate change and resource depletion remind us. For that reason, it is far more useful to think about why he was wrong rather than wallow in the "futility" of ever trying to think ahead, gratifying as some find this to be.

My assessment is that Kurzweil's predictions had two fundamental weaknesses. One is that he succumbed to "futurehype" about a few key technologies (in particular neural net-based pattern recognition), expecting that they would be much more advanced than they really were. The other is his facile view of the "softer" side of his subject, in particular politics, macroeconomics, and I daresay, ecology, given the role of oil supplies in our current troubles. Both are inextricable from the prevailing "wisdom" of the late ‘90s tech boom that was just about peaking when he was writing.

Of course, I should acknowledge that Richard Dooling, in his recent book Rapture for Geeks: When AI Outsmarts IQ, was more forgiving in his assessment of some of these same points. Additionally, one can argue that this is only the beginning of 2009, and that a truly fair assessment will have to wait for New Year's Day 2010; or failing this, that he might prove to have been just slightly off where some of his technological predictions are concerned, perhaps pointing to the recent proliferation of e-book readers.

Nonetheless, others seem to be clearly much further off, like the claims for telephone call translation and virtual reality. And the fact that so much of this change has come along much more slowly than he suggested is itself significant; the acceleration of technological evolution is the very foundation of the technological Singularity to which he has devoted so much ink, and with which he has become so closely identified.

No comments:

Subscribe Now: Feed Icon