Jan 202014

Dah-Jye Lee, an engineer from Brigham Young University, has developed a genetic algorithm which learns by itself to recognise arbitrary objects. This project is particularly relevant to the content on this website; a slowly developing augmented reality application still in research phase. I would predict that processing speed of the genetic algorithm is a bottleneck to performance with this method. The use of cloud processing and pre-populated data sets would bring this program nearer to real time functionality.

I will definitely be keeping an eye on further developments from this camp.

Source story here.

 Posted by at 5:45 pm
Dec 112013

The Google Brain can be accurately described as an industrial scale AI ¬†project. The first order of business for Google is to return the most relevant search results in response to a user’s query. The more naturally Google’s computer can “read” all the documents which it has indexed, the better the content it can deliver to this end.

So Google continues to build and develop systems which have to be more and more autonomously intelligent, to fulfill their business objective. Building these systems brings other benefits and capabilities, some of which may be even greater than the original purpose; a fact not lost on leading AI expert Ray Kurzweil. Ray has involved himself in the Google Brain project for the reason that they have very deep resources, which can lead to better quality and faster progress in such areas as understanding the workings of the brain. Knowing how brains work has huge advantages for humanity and the world in general.

See the source here.




 Posted by at 4:50 pm