Hacker News new | ask | show | jobs
by ekimekim 4192 days ago
I've often thought about how long until an entire system can be automated to the point where zero humans are involved - not in the running, maintenance, making money to hire the servers to run it...how would the law deal with such a system? How would society?

The closest analogy I can think of is legal trusts, that have a stated purpose with little room for interpretation, but still has a human performing the operations.

3 comments

We'd probably just deal with it the way we'd deal with wild animals and other things that ignore human laws. You might have some people with an interest in the legal rights of automated software just as some people have an interest in the legal rights of animals. But at the end of day, a dog can't appear in court and neither can a computer. So if humans decide to collectively work together to terminate a rogue piece of software (much like we already work together to try to stop malware), that's the end of it from the law's perspective.
What if it formed (or was) a corporation?
Someone has to own it and be responsible for it. Probably requires having a registered agent in some jurisdiction somewhere, for example.
Ying/yang corporations. Either corporation owns the entirety of the other.
I think Charles Stross(?) has featured this in books, with creations with bylaws that can be resolved as software instructions creating other corporations with bylaws resolved as software instructions, etc, with assorted corporations appointed as officers of others, creating tremendously deep layers of autonomous shell corporations.
Accelerando by Charles Stross is the book you're thinking of.
Thanks.
How do we find out, if not by building such a system? :)
If you want to find out more about this, you can google Autonomous Corporations. Some say that this is the real killer app of Bitcoin - because before BTC such things were quite impossible (there always had to be a human operating a bank account)