From movies to manifestos, few greatly hyped things ever live up to the promises made about them. The film of the book is always the film of the back-jacket blurb, and so on.
And, let's be honest, technology lets us down more than most things. Nuclear power was going to be clean, safe and too cheap to meter — instead of which it's a long-term health hazard, and a source of good stuff that terrorists want. Email was going to liberate us, instead of which it often seems to enslave us.
As the winter casts a chill on our souls, here's a run-down of the greatest disappointments we've had with technology.
1. Artificial Intelligence
What they promised...
"Machines will be capable, within 20 years, of doing any work a man can do," said Professor Herbert Simon, one of AI's founders, in 1965. Simon and others such as Marvin Minsky spent the 1950s and 1960s racing to use the computers of the time to simulate and potentially equal the functions of the human brain.
In 1968, HAL was refusing to open the pod bay doors in 2001: A Space Odyssey, and we were on the way to an exciting future where robotic intelligences would be our saviours, or even our eventual masters.
In subsequent years, AI research built "expert systems", and designed programs that learned and attempted to mimic the function of the brain with neural networks. The 1980s saw a huge boom in AI machines and languages such as Lisp, driven by fears that Japan's "Fifth Generation" project would be first to unleash the armies of intelligent machines.
... and what we got
In 1985, 20 years on from Professor Simon's confident assertion, what did we have? MS DOS. We even had to wait another 10 years for Mr Clippy.
The first AI boom fizzled out in the 1970s when the military agencies funding it realised the computers of the day couldn't handle the complexity involved. The 1980s "Fifth Generation" boom collapsed in 1987, and AI researchers went underground, using the gradually increasing processing power of systems to handle specific problems.
Now, pretty intelligent systems work on datamining, speech recognition and other applications. And, sometime in the future, we may indeed have to deal with really intelligent machines.
Meanwhile, since HAL, we've had a ton of AI in fiction, such as Steven Spielberg's 2001 film, Artificial Intelligence: AI. But the biggest public coup for real AI? The moment when IBM's Deep Blue beat Gary Kasparov at chess in 1997.
2. Ultrawideband wireless
What they promised...
The hype for ultrawideband is the finest kind of hype: it's true. UWB is a radical radio technology that really can send data at gigabit/s speeds over short distances. It's Bluetooth on steroids, if you like, and we expected it to be carrying high-definition TV signals around our homes by 2005.
It works by sending pulses of signal, at very low power, across a wide spectrum. You don't need a licence for that spectrum, because the power levels are lower than the amount of radio signals any device — your CD player, say — is allowed to leak.
The products will be fast enough to transfer a two-hour movie in 20 seconds, or a whole MP3 album in a second, without plugging anything in. It's also beautifully power-efficient, managing a massive number of bits per second on a small amount of power, making it perfect to use in small devices.
That ought to be enough to fuel a whole new business model — topping up our MP3 players and entertainment consoles on the go, with no fiddly wires or long waits.
... and what we got
So far, we've had pretty much nothing. First there was an endless wait for the standards bodies to agree on what version of the technology to use. Then there was a hiatus while regulators such as Ofcom wondered whether to allow the upstart, or to listen to the mobile operators who tried to convince them it would interfere with the phone systems they run in their precious licensed spectrum.
UWB had some good breaks. It was accepted as the next, faster version of Bluetooth, giving it access to potentially billions of phones. It's also been anointed as the wireless evolution of USB.
But despite all this, the silicon vendors have consistently over-promised and we're still waiting for products we were told would be...