Killer robots

[conflux]

I'm sitting here listening to a book on BBC internet radio recommended on UK poly, as it is to late for loud music. I have been working at home the last few days rather than drag myself in when I have a cold. This has meant I've been going to bed late and getting up late. Being able to work from home is a two edged sword. It is good to be able to avoid the commute but when ill I feel obliged to still do as much work as I am capable of, as so many of the people I work with still drag themselves in when they have colds.


I seem to have ended up writing comments about robots taking over the world a few times now. So I thought I would post a comment I made in vamp_ire's journal that was inspired by what she said about Blade Runner.

I think we have a very long way to go before robots will even get close to the kind of general intelligence that we are capable of. I would say that this was still 50 years away at least. If we ever reach this point then we need to code in the safe guards, from day one, to make sure that they want to help and work with humanity rather than anything else. Azimov's laws of robotics (or something a lot better) are currently a long way off but one day we may need them. The problem comes with the military. Building robots to kill is something that I think will one day need to be banned. Unfortunately, the US loves spending lots of tax dollars on subsidising technology research through military funding.

I also don't think nano robots will ever escape into the environment and do nasty things. This is because I don't think they will ever be able to operate independently of a large powerful controlling computer due to size constraints. We are already getting close to atomic scales in the latest processor designs and they are the size of a finger nail, so how would you fit that into a nano robot? The most complex microscopic thing is currently the human cell and they are not intelligent until they work together and are networked together in a tight formation. Possibly you could network the nano robots in this way but they wouldn't be able to sense and control all of themselves while in this mode, as all sensor inputs would be different. So you just get another large scale computer core with a small number of active nano robots controlled by it.

So in the blade runner film they have merged genetic research with robotics to create something that is becoming almost indistinguishable from a human being. They still built in safeguards, which highlight the lack of belief that they were becoming human, and which were inhumane as a result. Can you tell I'm a fan? Personally I think we will see genetic enhancement and cyber enhancement of mankind long before we would see robots becoming human because we can already make new humans in the same way as we have for millions of years but having the ability to upgrade ourselves like we can now upgrade our computer will become irresistible for some people. Just look at plastic surgery, and parents paying for their children's plastic surgery in certain parts of the US. Better to improve them before they are born? That individuals will be able to mess with the results of thousands of years of evolution is a very scary prospect.

Change is inevitable. We have to make sure that we choose the right path for that change. Frankenstein may be the story that saves the human race.

Comments

[ciphergoth.livejournal.com]

http://www.singinst.org/friendly/index.html

http://libertypunk.blogspot.com/2004_01_11_libertypunk_archive.html#107401377773551376

[lovingboth]

I'd be amazed if the US Military weren't spending hundreds of millions, if not billions, on killer robots.

Robocop (using a human brain instead) and Dark Star (the perils of having 'intelligent' computing so cheap that you can afford to put it in anything and everything) are my two favourite visions of a more likely future.