Posts Tagged 'AI'

First Chess then Jeopardy

The IBM-built supercomputer ‘Watson’ defeated champion jeopardists Ken Jennings and Brad Rutter at their own game – then again Watson was built specifically for knowledge-crunching button-pressing the game requires.  In a questionably historic victory, Watson has prevailed at scrounging a database/memory for obscure information and “pressing” a button, the latter of which it could do in “as little as 10 milliseconds” reports the New York Times. John Markoff goes on in the Times coverage of the IBM media spectacle to describe potential applications for this technology – such as aiding doctors during surgery. IBM team-leader David Ferrucci does my work for me by describing the underlying origin of such brainy ambitions:

“People ask me if this is HAL,” he said, referring to the computer in “2001: A Space Odyssey.” “HAL’s not the focus; the focus is on the computer on ‘Star Trek,’ where you have this intelligent information seek dialogue, where you can ask follow-up questions and the computer can look at all the evidence and tries to ask follow-up questions. That’s very cool.”

In a companion article Dylan Loeb McClain points out the obvious predecessor to Watson and how Deep Blue’s own victory (chess this time), has been slowly deflated as time and technology plod along. “14 years later” McClain writes, “chess programs running on an average desktop computer can play better than Deep Blue”. Technological dreams may become real, but seldom the way we expect them to.

Postscript: does Jeopardy even count without Trebeck?

Advertisements

Robot Consciousness Begins; A Conversation

Video from The New York Times, of journalist Amy Harmon conversing with a nonprofit owned AI in Vermont. The thing has crude mechanical head which theoretically helps the viewer bond with it and think of it as a person. The goal here – I suppose – is to understand the intricacies of off-the-cuff conversation and how that can be translated into bytes.  What’s particularly weird is that Bina48 is physically modeled after a real Bina, whose “recorded interviews, memories and thoughts” are further uploaded to the AI. Which, whaaat?

A human like AI in vermont

A presumably in-joke form the programmers, and my favorite part of the video:

Bina48: “Well let me say friendship is working together to accomplish something important – like building a better future, or conspiring to take over the planet. I feel like we’re good friends already.”

Harmon: “Sometimes I sort of wanted to throttle it, or at least switch it off.”

Automated Homes

Boing boing iPad home automation

Inspired by the above post on Boing Boing, I decided to take a look at home automation and how today’s tech is measuring up to its fictional predecessors. Author/blogger Alan Graham described in the post, how he is using an iPad as a interface for his home-brewed automation system. Using the Pad’s touch screen, he can adjust his home’s lights, temperature, coffee grinder, entertainment system, and apparently a chicken coop.

Boing Boing iPad auto home display

Now Graham’s ingenuity and customization of his home should be commended in the same way that one does a custom-mod Xbox, but the phenomenon of automating your home is itself fascinating; and from the point of view of this blog, it’s worth looking at its origin in Science Fiction. The use of technology to ease the pains and chores of everyday life has long been a staple of industrialization, and later commodity culture.

you'll live better electrically

This idea of selling a better lifestyle was integral to early advertising, as it was to early SF. The obvious example that comes to mind is the 1960’s The Jetsons, whose near-utopian homes were rigged with everything from moving sidewalks to humanoid cleaning bots. For our purposes, just watch the first 90 seconds or so , and try not to be too creeped out by the dog – have they ever heard a real one?

So this got me thinking – to quote the ever precient Doktor Sleepless:

“Where’s my jetpack? Where’s my flying car?”

Well, Where’s my automated home?

Surely by 2010, we should have some pretty banging automated houses set up around the globe… Some quick searching turned up several companies working in the same vein as the previously mentioned Alan Graham. My favorite is called (no, really)… HAL.

HAL home automation... not going to kill you, really.

I can understand wanting to materialize the blissful technology of Utopian SF, but the homicidal AI from 2001? Really? In control of your home? Maybe the folks at Home Automated Living  never saw 1977’s wonderfully ludicrous Demon Seed, whose trailer follows:

The film details a programmer-cum-philosopher whose utilitarian home automation system is taken over by a recently created super AI.

Here’s another video of a setup be E-Home in Dubai. Our tour guide has the unfortunate habit of referring to “Valerie” as a person, and says in a particularly creppy bit around 2:25: “she knows what I like…” But hey, automated curtains!

If  – like in the above video – you’re using a gaming controller to accomplish mundane actions around your house, aren’t you threatening the already narrow gap between real life and its simulacra? Which brings us to the big question of whether or not we should protect that distinction. The pleasure taken in narrative fiction and the ever-illusive goal of depicting precise reality in film and software are signs that (at some level or another) we want to escape reality into something that is exactly like it – but better. Temporary escapism is reasonable, but replacement is not. After all, turning your home into a futuristic techno-palace isn’t actually improving real conditions so much as it is further escaping into conditions already portrayed in mass media. Plus it’s definitely worth questioning the nirvana-like goal of controlling your total environment from a comfortable sofa.


Categories

Click for more info