Human in the loop

It would be great if we could just let automated agents take on complex tasks and trust that their output was clean and accurate so we can focus on the things we like doing. Unfortunately, this is far from the case, as the underlying logic of current artificial intelligence is inherently flawed with generative hallucinations to satisfy objectives. In a technology where the machines are meant to free human experience, we are left enslaved to it by having to check its work and constantly clean up its mess.