From: Phillip Huggan (email@example.com)
Date: Sun Jan 15 2006 - 16:46:01 MST
Well you need to get her drunk first. A massage. Good music helps too...
Seriously, I don't think it's wise to make conscious beings. There are ethical problems such as: if we piss it off it might kill all of us. I'm hoping AI will be done without minds.
I've learned through nanotech studies that deposition reaction procedures are very substrate specific. I would think this will carry over to minds too. If we want to clone Jeff Medina, I don't think any amount of silicon-based brain structures are going to do, no matter what topography we use. We would need to use carbon-based chemistries. There are EM fields that can be detected originating within human brains. Transposing carbon with silicon would change/eliminate the fields. The fields themselves aren't what is conscious, but for now things like an EEG are the best markers we have.
As far as I know you can't reproduce EM fields on conventional computer simulations. They're just there in the physical world. As soon as you produce a brain architecture that shows electrical activities and emits EM fields similiar to what our brains emit, then it is time to be scared for the human race. I don't know what kind of architectures this encompasses, but I know it rules out 2006 computer hardwares. There was a kid who made a Scanning Tunneling Microscope out of lego, but the fineprint reveals he used other non-lego components.
Jeff Medina <firstname.lastname@example.org> wrote:
On 1/15/06, Phillip Huggan wrote:
> Even if you can reproduce all the behaviours of the
> molecule, you still will probably need the actual atoms to play
This is a widely contested/unaccepted claim among computer scientists
and physical scientists, Phillip. Hence, I'd greatly appreciate if you
would include arguments for your claims about what probably will or
will not be needed in the future to create a conscious being.
Got holiday prints? See all the ways to get quality prints in your hands ASAP.
This archive was generated by hypermail 2.1.5 : Thu Jun 20 2013 - 04:00:34 MDT