 |
 |
Keith Ackermann
writes: He's getting older This is the first time I have seen him express intolerance. It is not uncommon that successful people become impatient with what they perceive as drivel as they get older. He has probably thought long and hard along many of the same avenues as those he dismisses. It becomes old after a while. What is craved is something new and riveting. He is immensely talented and I enjoy his work very much, but I do think he is wrong about a few things. The field of AI, in its present form, is laughable as a model for human thought. It's not a matter of scale either. The Turing Test is a joke too. You show me an AI that engages in hubris, maybe cruelly kills a dog for no good reason and then lays low, completely sated for a couple of weeks, and then we can discuss human AI. Until then, its all about designing wind-up toys. I think Kurtzwiel should take off from the ground first instead of jumping off a cliff. He should first build something 1/4cm long that flies around a person's head while avoiding being swatted, and then lands on the ceiling to emit a couple of larval self-replications before it recharges for the day by eating a tiny fleck of poop or whatever flies eat. Life is life, and I'm sorry, Doug, we may be able to whip it up in a blender some day, but we will never understand it. It's the reason the biosphere employs random mutation as a strategy to fill up all the nooks and crannies, and not some deterministic algorithm. I think Penrose, with his one-graviton threshold is a bit warner in the quest, and he stopped after the firing of a neuron. That's a lot of brainpower put toward a most tentative supposition of the non-determinism that might be involved in firing a single neuron. |
| [358] Posted on Thursday, 25 December 2008 at 2:44 GMT [Reply to this] [Permalink] |
©1997-2022 by Tal Cohen, all rights reserved. [About]