Breaking News
Blog

Paul Allen Makes Oren Etzioni New AI Chief

Carolyn Mathas
9/18/2013 00:00 AM EDT

 2 comments   post a comment
NO RATINGS
View Comments: Newest First | Oldest First | Threaded View
CMathas
User Rank
Author
Re: What if they succeed, and create a human-level AI?
CMathas   9/20/2013 11:27:28 AM
NO RATINGS
I tend to agree with you as to it will be very good for us or disastrous. I'm sure you remember Bill Joy's article in Wired: Why the future doesn't need us.

http://www.wired.com/wired/archive/8.04/joy.html

I watch with awe at what is being done in this area of AI, yet there are so many ethical issues involved. 2050 seems like it's way too far out, in my opinion. I'd be surprised if it took more than half that time.


Thanks for your comment--what do the rest of you think?

Calum
User Rank
Author
What if they succeed, and create a human-level AI?
Calum   9/19/2013 1:50:50 PM
NO RATINGS
Paul Allen has not announced any targets for what AI2 will produce and by when, nor how much money he will pump into it, but he has already provided $400m for his brain research institute, so it won't be peanuts.

€1bn for Henry Markram's Human Brain Project, $300m and counting for Obama's BRAIN (Brain Research through Advancing Innovative Neurotechnologies initiative, plus whatever Google and IBM are is spending.  Pretty soon we'll be talking real money.

And what is this money being spent on?  The creation of a human-level AI.  We cannot know for sure yet whether and when this project will succeed, but many experts expect the first human-level AI to arrive before 2050 - possibly much sooner. 

Human-level AI will either be very good for us - or disastrous.  Too little thought is going into this.  A possible roadmap  is described here: http://bit.ly/170uDbM.

Most Recent Comments
michigan0
 
SteveHarris0
 
realjjj
 
SteveHarris0
 
SteveHarris0
 
VicVat
 
Les_Slater
 
SSDWEM
 
witeken
Most Recent Messages
9/25/2016
4:48:30 PM
michigan0 Sang Kim First, 28nm bulk is in volume manufacturing for several years by the major semiconductor companies but not 28nm FDSOI today yet. Why not? Simply because unlike 28nm bulk the LDD(Lightly Doped Drain) to minimize hot carrier generation can't be implemented in 28nm FDSOI. Furthermore, hot carrier reliability becomes worse with scaling, That is the major reason why 28nm FDSOI is not manufacturable today and will not be. Second, how can you suppress the leakage currents from such ultra short 7nm due to the short channel effects? How thin SOI thickness is required to prevent punch-through of un-dopped 7nm FDSOI? Possibly less than 4nm. Depositing such an ultra thin film less then 4nm filum uniformly and reliably over 12" wafers at the manufacturing line is extremely difficult or not even manufacturable. If not manufacturable, the 7nm FDSOI debate is over!Third, what happens when hot carriers are generated near the drain at normal operation of 7nm FDSOI? Electrons go to the positively biased drain with no harm but where the holes to go? The holes can't go to the substrate because of the thin BOX layer. Some holes may become trapped at the BOX layer causing Vt shift. However, the vast majority of holes drift through the the un-dopped SOI channel toward the N+Source,...

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
Like Us on Facebook
EE Times on Twitter
EE Times Twitter Feed