Singularly speculative

Posted on November 28, 2011


Dateline: 12 October 2011 // Posted by: AHJOConnell
Source: Forbes

Story: “Ray Kurzweil On The Future Of Innovation At Singularity University”

Summary: The article tells of Ray Kurzweil speaking to a group of his students at the singularity university in Silicon Valley, California. He is talking about the great strides that have been made in technology in the last few years but most importantly the strides that will take place in the not so distant future.

The article mentions many powerful people in the audience watching Kurzweil’s speech such as CEOs, representatives from IBM, and even the Brazilian military. Kurzweil, himself is not actually physically attending the lecture, merely addressing the assembled crowd via teleconference appearing (or as the author so brilliantly put it, “onscreen like Oz the great and powerful in High Def“).

Kurzweil spends a great deal of time explaining how technology is progressing at the rate he expected according to his projections. Though not stated in this article Kurzweil has for a long time predicted that singularity — the moment at which artificial intelligence will surpass human intelligence — will have occurred by 2045. Also, not stated in this article is the fact that Kurzweil is often accused of adjusting his projections every year to back-up his case.

Why is this PseudoNews? Kurzweil appears to make several assumptions about the future that, unsurprisingly, cannot be proven right now. No one can know whether Kurzweil’s claims regarding 2045 are true or false. As such, such claims cannot be said to be supported by empirical evidence in the normal sense of of the term.

Also, Kurzweil speaks of a device the “size of a blood cell”  and “a billion times” more powerful than his smart phone being a reality in 25 years. Again, there is absolutely no way to test this and while Kurzweil may claim that he takes many factors into consideration, he fails to account for many social factors such as the possibility of there being no demand for something so small. As such, to put such a specific date on it is downright foolhardy. Also, the use of terms such as “billion” seems to be very vague: for example, he states that his smartphone is “a billion times more powerful per constant dollar” than the computer he used in MIT as a student. Now there may well be some economic formula to corroborate this, but it in this context it appears more likely to be presented purely for effect and to make the super powerful blood cell computers 25 years down the line seem like a greater reality. “A billion times greater” is language that sounds more like something a child would say while arguing in the schoolyard than a scientific measure. It is a very vague number that is impossible for most people to imagine.

The author of the Forbes article acknowledges that Kurzweil has his critics, who describe him as “too far-thinking”. Nonetheless, the article itself does not challenge Kurzweil’s most outlandish claims, most notably about singularity, and seems to be broadly supportive of his teachings and philosophy.

What features of pseudoscience are on show? One of the main features of pseudoscience here is that Kurzweil’s predictions are unfalsifiable. They are merely predictions based on past trends but are more or less presented as scientific facts by Kurzweil.

Another feature of pseudoscience is the idea of over-reliance on a guru figure. Kurzweil’s students have paid fees of around $25,000 to  attend his summer tuition programme. To his fans, Kurzweil is something of a “go-to guy” in regards to anything related to singularity and his predictions are seen as near-gospel. Pseudosciences are often characterized by focusing mainly on the work of one main individual and this is very apparent in this case here.

There appears to be a risk of confirmation bias. Kurzweil is giving his speech to a group of people who have spent large sums simply to be in attendance, which may sway the majority of his audience to believe teachings and, indeed, to perceive anything he would say in positive terms.

There is also an element of science by press conference given that, while Kurzweil has had many books published, attempts to locate peer-reviewed journal articles by him on the subject often appear futile. Books do not tend to be professionally peer-reviewed in the same way as journals, and public lectures certainly aren’t. Given his dissemination strategy, Kurzweil seems more concerned with getting his message out to people without scientific backgrounds, than to more scientifically minded individuals who would perhaps be a lot more skeptical.

Finally, there is of course a lack of parsimony as Kurzweil’s claims are based on several unproven assumptions. Parsimony means that a scientific theory should make as few new assumptions as possible seeing as Kurzweil’s theory is nothing but assumptions, to describe it as parsimonious would be borderline impossible.

The moral of the story: While Kurzweil’s predictions might eventually prove to be correct, they rest on much the same basis as those of a gambler placing bets on a horse race after studying past form. His measurements are not scientific but are presented as if they are and as a result can be said to be an example of pseudoscience.