Some morning musings on AI. We need to stop talking about this as technology issue. It's an ethics & law enforcement issue. Thread:
5
1 Boost
0 Likes
2. So let's instead come from an ethics angle. Broadly speaking, companies and individuals can operate at 3 ethical levels. At the top: do what's ethical. Slightly below that "do what's legal". And below that "do what is not cost-effectively-prosecutable"
1
3. When companies or individuals operate anywhere between that which is ethical and that which is legal, the government is a partner in their success. When they operate between what's legal and what's prosecutable, we have a law-enforcement problem.
1
4. Now, as a legal matter, crimes are defined by actions, and by motive. They aren't defined by weapons. Murder charges don't vary depending on whether you use a gun, or a slow, Agatha Christie poisoning. The "technology" doesn't limit the legality.
1
5. So now transplant all of what's happened in AI lately to any other industry. Suppose a pharma company releases a product that millions of people find useful. They do de minimus testing prior to release. Then they say "turns out this product might kill you, but I can't stop making it."
1. Framing this as a technology issue leads us down all sorts of unproductive rabbit holes that only stall action. "we don't really know yet how this will play out", "don't get in the way of innovation", "we have to outrace the competition", etc.
2/2