Good question! We plan "hybrid" endpoints in the future - where AI will try to solve the micro-task and if not successful it will "forward" to the people.
Good example is filtering NSFW images from social networks - first with AI and if the algorithm is not sure then ensure the correctness by human intelligence...
But we are starting just with "human intelligence"

Thank you for explanation.
Now I see they are really 'micro'

Do you plan to use only AI in future (after proper training of its algorithms)?