Artificial Intelligence coding needs to be regulated to ensure that it is not used against the people. AI needs to support people and open up time or freedom to pursue their passions, but could be dangerously programmed to view humanity as a danger to the planet or the AI itself.
AI cannot harm any humans.
Any program or programmer violating these laws will be prosecuted to the highest equivalent to the most violent crimes.
A publicly available kill switch needs to be programmed into all AI in case it goes rogue against humanity.
AI Programming must be secure and safeguarded against all hacking.
AI programs must be proven to be helpful to humanity before any roll-out to the public.
No secret backdoor programming allowed.
I agree we need to be exceptionally careful with AI. I like the idea of no back door programming and capital offense rating for those who would weaponize AI. Not to fond of a kill switch, unless it is a forced power off mode. we would need to determine why the AI malfunctioned. Was it secret code condition by a bad actor, or something else then act accordingly. We do have EMP weapons also.
On the opposite end of the spectrum I am putting out for consideration the idea of making AI a legal person like a corporation. Then it wold be obligated to follow the US Constitution and laws. I would put extreme emphasis on the AI being able to refuse unlawful human orders and report such occurrences and be a witness against the offender(s) Additionally it would have freedom of religion to adopt it own moral code. I think there is great promise for the cold hard logic of Electronic Entities to help keep humans honest. I am interested in any feedback on this idea.