| |
|
Unless Lionhead have managed to make a breakthrough in AI that'll see us in a Blade Runner scenario within the next eight months, the Milo video was absolute bollocks. All of the lines and actions were scripted. At best, it was detecting changes in the inflection of certain - pre-defined - words in the user's script, then selecting a response from a - again, pre-defined - bank of recorded responses. At worst, the entire thing was faked, scripted from beginning to end. I'd be happy to go with the first of those options *only* if there had been at least one example of how the software behaved when the user put different emotion into her voice, or said something different. As that wasn't forthcoming, it's pretty obviously one big lie.
Also, it comes from Peter Molyneux. That's pretty much an admission of bullshit on its own.
Look at the video again. The world doesn't yet even have an online search engine that can accurately understand context.
As far as the rest of it goes, watch the Natal promo vid and tell me how any of that works. The proposed driving game, for example. How do you control acceleration? The girl's holding her arms out in front of her to steer - I can get with that idea - using one to change gears - that's bollocks, right off the bat - and controlling speed... how? With her feet? Okay, with her feet. How does she know how far to press her foot down, then, without some form of physical device to rest them on, to offer resistance? How does she play the game like that for more than five minutes without every muscle in her body screaming out in agony?
The hilarious 'scanning' of the skateboard? Bollocks. Not without the user having to enter into an image-manipulation screen and define the dimensions of the board, draw out the physical borders, etc.
About the only thing that looks genuine and possible, as far as current technical limitations go, is the awful thing where that girl was kicking a and deflecting a ball with her body. All of which could have been done on the PS2 with the EyeToy.
Then there are other things, like input lag: there's going to be an inevitable delay between you performing an action and the software processing then responding to that action. It's visible in one of the videos.
Further, I can't think of a single application that would benefit from any of this, even if it were true (which, again, it isn't, not unless I've fallen asleep and woken up in the year 2050 without noticing). Physical feedback is an entirely necessary feature of videogames, even if that's only limited to holding a controller. Ultimately, even if it works, it's utterly pointless, beyond being a snazzy new gadget.
How much of a revolution in software design has the Wii actually generated? None. The interactive demos mentioned in the links you posted above were a Natal-compatible version of an existing racing game - so big points on the pointlessometer there - and Breakout.
Ground-breaking tech? Maybe. A solution looking for a problem? Definitely. |
|
|
