Two thoughts:
1. If children in their early years are better at reconstructing language grammar on the fly than adults, is it possible to come up with a different task that their brain could solve better than an adult’s brain, if its formulation would mimic language?
2. Is it possible to devise a practical task (not a CAPTCHA) that, when broken down into many elementary components, could be solved more effectively by a group of specially trained/prepared people than by a modern computer?
Hi, I'm Rauf Aliev. 