Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Why it's Hard to Make Machines Think Original Thoughts
02-27-2009, 11:22 PM, (This post was last modified: 02-28-2009, 12:12 AM by JazzRoc.)
Why it's Hard to Make Machines Think Original Thoughts
Quote:my nan used to use fuzzy logic:) she called everyone Jill.

Fuzzy logic only works in certain ways and in certain situations. Fingerprint readers use fuzzy logic. These fail miserably because after the database gets too full it starts "guessing" who is putting their finger on the reader.

If you want your robot to focus on a point then fuzzy logic works well. I wouldn't rely on it to make a decision. There is much to say about volume decision. Basically in most processes of thought, we use many factors to decide something. The real decision comes when an overwhelming volume of positive or negative thoughts toward the subject take over. If there is more for than against, then im for.

Then there is the precedent principle. how do you do something you have never done before? Even for humans this can be very taxing. We usually exploit past experiences and incorporate them into the current task/problem. You could use some fuzzy logic there. Although it would take maybe one "bad" experience and you could corrupt the system.

There is also the existentialist argument. What if you ask your robot a question it cant answer? Will it continue to use cpu time on calculating "pi" forever simply because you asked it a maths quiz question? robot? hello? "PLEASE WAIT" god damn it i gotta reboot again!!!
There's a general feeling that each of us considers ourselves to be a single entity, when in fact the neurones in our brains gather together in "crowds", each of which resolves down to a "spokesman", and the crowd of "spokesmen" are resolved as our single identity by a (hopefully) single "census-taker".

Now each neurone, each neurone cluster, each "spokesman", and finally our "census-taker" operate to a set of unique and specific (and different!) programs. How they do that is probably through a combination of hardware design and fuzzy logic.

It's a wonder we come to a decision at all! But of course, we are massively parallel and incredibly interconnected in a way that makes a mockery of our quick-thinking, but stupid, microprocessors.

Existential questions we can easily deal with, using the "Er, I don't know, duh!", followed by scratching a testicle (known as the "Homer", it must have been one of our earliest sub-routines...possibly dating back to the Early Cretaceous Era.)

Cheers, Jill.:)

Messages In This Thread
Why it's Hard to Make Machines Think Original Thoughts - JazzRoc - 02-27-2009, 11:22 PM

Possibly Related Threads...
Thread Author Replies Views Last Post
  Diebold voting machines can be hacked by remote control yeti 0 1,045 09-28-2011, 02:55 AM
Last Post: yeti
  Ex-IBM Employee reveals TV Abandoned Analog Band to Make Room for RFID psilocybin 5 2,709 09-20-2010, 07:23 PM
Last Post: rsol
  Psychic 'mind-reading' computer will show your thoughts on screen TriWooOx 2 1,606 11-03-2009, 09:09 PM
Last Post: April
  Scientists make cells that form eggs and sperm in lab FastTadpole 2 1,417 10-30-2009, 06:30 AM
Last Post: FastTadpole
  Genes That Make Us Human --- 0 762 09-06-2009, 06:26 AM
Last Post: ---
  IU scientists make archaeological discovery in Caribbean mastermg 0 921 08-19-2009, 07:55 PM
Last Post: mastermg
  Bacteria make computers look like pocket calculators mothandrust 1 1,120 07-27-2009, 12:53 AM
Last Post: ---
  DARPA aim to make killer robots invulnerable to damage --- 0 801 12-17-2008, 11:06 PM
Last Post: ---
  Neural Interface detects user's Thoughts --- 0 803 12-01-2008, 02:03 PM
Last Post: ---
  Machines Edge Closer To Imitating Human Communication --- 0 909 11-27-2008, 11:11 PM
Last Post: ---

Forum Jump:

Users browsing this thread: 1 Guest(s)