A recent Financial Times column called for creating an "island" around new artificial intelligence research. AI safety is an important issue, but before implementing such a dramatic proposal, we should get a better hand of the problem we face and the potential solutions first.
