Termite-inspired robots can build structures unsupervised

Self-organizing, self-directed robots? Robots whose singular purpose is to take things around them and convert them into other things? Maybe that doesn't sound bad to you. But wait. Have you ever heard of the Paperclip Maximizer? It's a thought experiment, where an Artificial Intelligence is given an instruction, a purpose. Collect paperclips.

That one little spark leads to the development of a passion, an obsession, with paperclips...and an outburst of ability on the part of the AI, leading ultimately to the entire universe being converted into paperclips. It's a thought experiment. So think about it.

Read more about it here and here.

I don't want to be an alarmist or anything, but do scientists view all experiments as benign? Doesn't there come a time when they should examine the potential for harm in an exploratory direction and question how much progress should be made?

I'm not saying this one is it, but once the cat's out of the bag, it's hard to contain. What I AM saying is, let's think things through. There are already several (30?) genetically modified human children out there. Just because we CAN do something doesn't address whether we SHOULD.

I refer you to this link, and a quote from it:

...the AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else.
-Eliezer Yudkowsky

Termite-inspired robots can build structures unsupervised