> But right now we have 0 clue how this happens in humans so we are not able to program it in computers. Without taking a position on whether or not computers

> But right now we have 0 clue how this happens in humans so we are not able to program it in computers. Without taking a position on whether or not computers can create new exploratory knowledge (how could one begin to say authoritatively that they can't?), this "so" doesn't follow. The fact that we don't know how to do something doesn't mean we can't program it in computers.