That's a bit far fetched.
As I understand it, successful integration of those things into computer chips would give us a new form of non-volatile memory. It would essentially do the same as flash memory today, maybe cheaper and faster but not with spectacular new functions.
And it would certainly not solve the little research problems that are still between us and serious AI.
Electronics-based AI breakthrough?
-
rabiator
- Posts: 92
- Joined: Mon Jul 26, 2004 11:51 am
- Location: ??
- Contact:
-
xerxes31415
- Moderator
- Posts: 305
- Joined: Sat Jun 16, 2007 12:50 pm
- Technosexuality: Built
- Identification: Android
- Gender: Male
- Location: Los Angeles, CA
- x 2
- Contact:
Yeah, AI is mainly a software problem. The software guys just need to really do some research and move it along. Then it's up to hardware guys (like me) to build the chips to make those algorithms and laws run in real time.
The memristor is cool, but I doubt we'll see any real effects for a few years, and even then, minimal.
The memristor is cool, but I doubt we'll see any real effects for a few years, and even then, minimal.
Users browsing this forum: No registered users and 11 guests