Robots

Here is an interesting story about learning from flies and simplifying the computational requirements of sight in order to better emulate vision and provide for a much lower cost way to detect motion and obstacles the coming generations of small robots.

http://www.wired.com/wiredscience/2009/11/fly-eyes/

I was just in a conversation (or is IRC only ever a fight) this morning where I was arguing against trying to construct large exact models of things in order to replicate human biological behaviours without looking around at simpler implementations that could be the inspiration or answer.

One of the upsides of having little children and insomnia is that I get some extra time to think about things I have read.

This article http://www.technologyreview.com/biomedicine/23354/page1/ by Ed Boyden of MIT was rolling around in my head the other night and led to some thoughts.

He touches on the futurist concept of singularity, a time of such rapid technology change that we cannot envision the future beyond it.

I am not a big fan of singularity as I know it today, there are a couple things that bother me

- The expectation that something like this should happen in or near this time period. This sounds like an 'it's different now, we are special' argument to me.

- Why such a limited focus and frame, on the grand scale of the universe, perhaps the last 40,000 years have been the singularity on this planet.

- And why does it have to include advanced AI, is the singularity just simply the day the machines take over, if so, then the human spec on the universe will be even shorter than we hoped.

The other topic he gets to is artificial intelligence needing a motivation and goals so it has a feedback loop to continue its advancement. Again this raises some questions for me

- I see really only two kinds of motivation, Ego (greed,want,recognition) and Primal (needs,death). So, assuming we are only smart enough to start our machine with one of these and no balance, take your pick.

- And why are we so interested in the idea of a wanting, self advancing machine? an army of Datas (ST:TNG) to do our bidding? I thought we answered that in the 80s.

- If we do make it here, the future I fear is not that the machine gets bored or sad or gives up, it is that we do. Once we build something smarter than ourselves and switch to sloth, perhaps we wouldn't know a singularity and that the machines had taken over if it happened to us anyway.

Enjoy the rest of your day inside the Matrix.