If social bots use APIs, wouldn't it be easy to ban them? Can't Twitter or Facebook easily tell the difference between malicious bots that use the APIs and users that interact directly with the UI in their site? And how they manage to create so many accounts to spam?
If social bots emulate browsers, wouldn't it be way too hard to maintain? The UI changes quite often.
What exactly is the strategy behind those spam bots that appears every time in Youtube, Facebook, Amazon etc?