WIRED:   PEOPLE IMPRINT on annual events, like cute chicks imprinting on unlikely predators in Warner Bros cartoons. They follow them around, long after they've learnt that perhaps they're not the mother lode of all good ideas, writes Danny O'Brien.

Half of San Francisco, both geek and defiantly non-geek, feels the same about the Burning Man Festival, a sort of survivalist Glastonbury in the Nevada desert.

Plenty of musicians and technologists are now sharing the same excitement at a conference in Austin called South By South West (SXSW). I'm not there, so I'm going to pointedly ignore them, although you can read their over-excited reports on much of the technologists' blogosphere.

If there was an event that I have imprinted on, it's probably tech publisher Tim O'Reilly's Emerging Technologies conference, which happened last week. I've gatecrashed it, run nearby shanty towns for impoverished hackers attending it, spoken at it, run competing conferences 500 miles away from it and, I admit, dined out journalistically on tidbits from it since 2001.

Emerging Tech is a conference where O'Reilly attempts to glean what the most mole-like and blinking of alpha geeks are planning. For O'Reilly, it's turned up the pearls that he can turn into other conferences and best-selling tech books.

If you've ever been driven crazy by the hype of Web 2.0, blame Emerging Tech - O'Reilly picked up many of the ideas that were to turn into that mini-boom from geek-watching Etech.

But the problem with any imprinted conference is that it has its time in the sun, then becomes slowly a dim echo of itself. The golden age is never quite decided upon - was Glastonbury's golden age the Seventies? Was Wexford all over by 1966?

Emerging Tech's difficult year was last year. Essentially, all the tech that had excited me and others had already emerged - Flickr, Delicious, AJAX sites, Web services, crowdsourcing and other Web 2.0 buzzwords were all being rolled out by the smart and the not-so-smart.

The alpha geeks who had to put up in my back garden were now earning gazillions implementing the crazy ideas they had yaddered about in the fallow years before and had no time to attend conferences, let alone speak at them.

But Etech had a small renaissance this year, by determinedly chasing those geeks past their old (but profitable) former projects to what is driving them now. For anyone in the Valley, the obsessions seemed straightforward, if . . . well, a little crazy. Space exploration; robotics; crunching vast data-sets; body modification; cognitive-enhancing drugs; re-engineering government.

You'll no doubt hear more of all of these as I chase down the ramifications of each in the next few months. What was notable about them all (and spotlit by O'Reilly in his keynote) was that none of these topics was particularly driven by a search for profit.

Perhaps only Google's talk on how it mines its huge archive of textual data for patterns had some obvious revenue potential, and Google's director of research, Peter Norvig, spent very little time talking about that.

Among this audience, he would rather just geek out on the applications - such as deducing the translations of webpages by comparing them with billions of other foreign-language texts that Google has: like using the whole of the internet as your Rosetta Stone.

Norvig is an artificial intelligence expert by trade. He sees the huge corpus of data that Google can gather, and is gathering, as steps to reach some of the lower hanging fruit of AI - machine translation, some level of comprehension by computers of the structure of our world.

I find the monopolistic and privacy aspects of that pursuit disturbing.

Answering a question from a member of the audience, Norvig conceded that no start-up would be able to catch up with Google in the data-collection market.

Nathan Eagle, another geek speaker, seemed even more amoral, talking about the behaviour and intentions he could deduce about individuals merely by observing their mobile phone manoeuvring and activity. Friends; enemies; who likes to hit snooze; who doesn't want someone else to know that they hate them: all was revealed in the data Eagle collected from mobile phone companies.

Not here, I'd hasten to add. No, Eagle went to east Africa to get his data, where phone companies are a lot more easily persuaded to hand over their customers' data to strange researchers.

If that sounds a little depressingly amoral for these alpha-geeks, let me reassure you that some, at least, have a sense for the wider implications of their work. By far the most compelling presentation of the conference was that by Saul Griffiths, a recent winner of the eclectic McArthur Genius Grant and rogue engineer of amazing devices for the military and the wider world - and also a children's book author (see his site, Howtoons, for a glimpse of his pedological bent).

Griffiths has kicked around the O'Reilly circuit for some time (pausing only to marry Arwen, O'Reilly's daughter), but now his speeches have turned serious.

He and others have abandoned their more light-hearted projects and have decided to cure global warming.

He's not sure how, but as he deconstructs and reconstructs the predictions and the possibilities like a young Freeman Dyson, you begin to feel more optimistic about all our futures. Not every Golden Age ends with a Dark one.