The city of Bloomington, Indiana, USA, has banned an AI that does not yet exist.
On November 4, the City Council voted unanimously to pass a local law banning the distribution or deployment of “supersmartphones” in the city. The legal definition is: Cognitive performance that significantly exceeds that of humans in almost all areas and the ability to act autonomously, such as rewriting its own code or operating without constant human supervision. Violators may face fines and court bans.
The sponsor is City Councilman Dave Rollo. He cited some researchers' estimates that the probability of human beings encountering a disaster is more than 50%, with one putting it at more than 99%.
The first draft was much more radical than the final version. The scope of the original version may be extended to residents who use such AI at home; to lift the ban, more than 95% of all registered voters will need to vote in favor. Both items were removed by amendment before the vote.
What can a ban in a small town control? Basically nothing can be controlled, and no company will deploy super intelligence in Bloomington.
Rollo himself knows that his memorandum is about another calculation: there are about 350 cities in the United States that are the same size or larger than Bloomington. If they all follow suit, super-smart companies will face "huge regulatory burdens."
So this is more of a demonstration, using local legislation to send a message to the state and federal levels. Whether there is a second city to follow will be known in the next few months.