Google Is Trying To Create A Human Brain Using 16,000 Computers

• Lan wrote on 2012-06-26 22:12
  [video=youtube;Ppv5qCYHSN8]http://www.youtube.com/watch?v=Ppv5qCYHSN8&feature=g-u-u[/video]
  

  Here's the latest project from Google's ultra-secret Google X Labs, where Google works on fringe technology projects like self-driving cars and computerized glasses: learning how to find cats in YouTube videos.
  Google X is developing a huge network of computers that are able to learn on their own what is present in a YouTube video through a process called machine learning, John Markoff at the New York Times reports.
  The Google research team, led by a Stanford University computer scientist named Andrew Ng and Google fellow Jeff Dean, networked 16,000 computers and fed them random thumbnail images, each one extracted from 10 million videos on YouTube, the New York Times reports.
  The research began years ago, in an attempt to simulate the human brain, the Times reports.
  By culling millions of YouTube videos and scanning thumbnail images from the videos, the computer cluster was able to determine what a "cat" is on its own, rather than being programmed to find cats in those videos.
  Here was the scientists' reaction, according to the New York Times:
  Currently much commercial machine vision technology is done by having humans “supervise” the learning process by labeling specific features. In the Google research, the machine was given no help in identifying features.
  “The idea is that instead of having teams of researchers trying to find out how to find edges, you instead throw a ton of data at the algorithm and you let the data speak and have the software automatically learn from the data,” Dr. Ng said.
  “We never told it during the training, ‘This is a cat,’ ” said Dr. Dean, who originally helped Google design the software that lets it easily break programs into many tasks that can be computed simultaneously. “It basically invented the concept of a cat. We probably have other ones that are side views of cats.”
  The Google brain assembled a dreamlike digital image of a cat by employing a hierarchy of memory locations to successively cull out general features after being exposed to millions of images. The scientists said, however, that it appeared they had developed a cybernetic cousin to what takes place in the brain’s visual cortex.

  
  
  http://www.businessinsider.com/google-uses-16000-computers-to-find-cats-on-youtube-2012-6
• Sumpfkraut wrote on 2012-06-26 22:15
  Nothing new, has been done before.
• Osayidan wrote on 2012-06-26 22:36

  Quote from Sumpfkraut;894946:

  Nothing new, has been done before.

  
  
  Usually by giving data to the machine so it knows what is what. This is a poorly written version of the article I read this morning.
  It mentions that no data was given to this computer about what is what. It can find cats (and other things) but it doesn't know what they are in human terms. It created its own idea of what a cat is, from seeing millions of images of them.
• ♪ wrote on 2012-06-26 22:38
  Yeah, this has been done a few times.
  Nothing new, and nothing amazing.
• Bakuryu wrote on 2012-06-26 22:41
  Nothing amazing?
  Do you even realize what is going on here?
  
  I just hope my sarcasm detection system is failing and your post was just that.
• Sumpfkraut wrote on 2012-06-26 22:45

  Quote from Osayidan;894958:

  Usually by giving data to the machine so it knows what is what. This is a poorly written version of the article I read this morning.
  It mentions that no data was given to this computer about what is what. It can find cats (and other things) but it doesn't know what they are in human terms. It created its own idea of what a cat is, from seeing millions of images of them.

  
  I'm thinking specifically about an experiment where they just put computers together, and noticed patterns of electrical current emerging by their own and that's it. Might be fake, I just read about it in a very small article a long while ago. I think it was relatively minor in scale anyway, about the activity level of some part of a rat brain (that was the example used). I can't find anything about it with a quick Google search though.
  
  

  Quote from ♪;894960:

  Yeah, this has been done a few times.
  Nothing new, and nothing amazing.

  
  Well it is pretty cool. But it's definitely not an outstanding concept.
• Chiyuri wrote on 2012-06-26 22:59
  the machine probably made a link with the word "cat" writen in the tittle of many youtube video containnig cat and the existance of what a cat it.
• Osayidan wrote on 2012-06-27 02:28
  The main idea is that the computer created its own concept of what a cat is, without any human intervention. It doesn't use the word "cat" (or any word from what I saw), it just developed a behavior where it can identify cats.
  
  It's like if you put a newborn baby with no human interaction into a room with a bunch of objects and a few cats (assuming it doesn't need to be taken care of to live) it would over time develop a concept of what the cats are, without giving them a name. Put the baby into another room with different cats and it would know that those animals are part of the group of whatever it decided cats are.
  
  This is a fucking huge advancement for artificial intelligence.
• Xemnas wrote on 2012-06-28 04:34
  It begins.(Sorry Osay you won't be the start of it)
• Drizzit wrote on 2012-06-29 04:47
  That's still a long way off from what an actual human brain can do. Humans ftw.