Sponsor's Message

These are always an entertaining read for anyone with an interest in technology, and there are numerous little gems buried within their pages.

For example -- did you know that in 1968 they invented a wet-suit which included an in-built heating system powered by a radioactive isotope? How nice.

There's even a feature on hydrogen fuel cells from the mid-60's which claims that "we may have to wait as long as a decade before these cells become commonplace in the cars we drive." Oh yeah -- they nailed that one right?

Regular readers of these magazines are probably sorely disappointed right now. After all, if PM and PS were to be believed, we should have landed on Mars nearly a decade ago, all have flying cars in the garage, and be able to enjoy holidays on the moon.

Readers Say (updated irregularly) Nothing Yet... Have Your Say

Okay, so the tech-writers and scientists of the 1950s and 1960s got it very wrong in respect to space exploration and transport, wildly over-estimating the rate of technological advancement. But they were also equally guilty of underestimating the pace at which computer technology would improve.

Even your desktop PC is orders of magnitude more powerful than anything predicted by the scientists of 40 years ago -- but that raises another question: why hasn't software evolved at the same pace as hardware?

Yes, we do have some nice fancy GUIs and some wonderful applications but some aspects of computer software are just like predictions of flying cars and moon-bases.

I refer to basic things like artificial intelligence and reliable voice recognition.

Even highly sophisticated computer systems such Deep Junior, which recently began doing battle with chess-master Gary Kasparov, rely not on any inate intelligence but crude, raw, brute force. It does nothing that couldn't also be done by a decade-old PC (albeit far more slowly).

And voice recognition? Well there are some systems that can be trained to recognise spoken commands with stunning accuracy -- but once again it's a brute-force approach with significant limitations. Many of these systems must still be trained to recognise an individual's voice, have limited vocabulary and still fail to work well with regular "fluent" speech.

You only have to watch TV presenters struggling to demonstrate one of the new crop of hi-tech cars to recognise their voice to see what I mean. Eight times out of ten their carefully spoken commands are recognised but all too often several spoken attempts have to be made to do something as trivial as change the radio station. And heaven help anyone with a lisp!

Will the next real breakthrough in computer technology be the development of technology that really does produce intelligence from a machine? Imagine how much of a leap-forward this would be.

What if we could integrate near-perfect speech recognition with an intelligent data-processing system? You'd have an instant multi-lingual translation device that would enable you to speak with anyone on the planet. Such a system could even be integrated into your cellphone, automatically recognising the language being spoken by the other party and performing transparent two-way verbal translation.

Unfortunately, the best we can do right now are things like the numerous online translators you'll find on the Web which have very little appreciation for grammar or the finer meanings of language.

Of course the downside of true artificial intelligence would be that a mountain of ethical considerations would suddenly emerge. Is an artificially intelligent machine actually a life-form? Will it develop emotions? Should it have its own set of rights?

The future may be more interesting and frightening than the publishers of Popular Science could ever have imagined.

Contacting Aardvark
As always, readers are invited to submit their comments on material covered in this column. If you'd like your comments published here then please be sure to use this form and select For Publication.

Other media organisations seeking more information or republication rights are also invited to contact me.

Just add a couple of lines of JavaScript to your pages and you can get a free summary of Aardvark's daily commentary -- automatically updated each and every week-day.

Aardvark also makes a summary of this daily column available via XML using the RSS format. More details can be found here.

Contact me if you decide to use either of these feeds and have any problems.

Linking Policy
Want to link to this site? Check out Aardvark's Linking Policy.

Did you tell someone else about Aardvark today? If not then do it now!