Share a tractable AI safety problem with the community and optionally attach a bounty
What type of problem are you submitting?