Why can't we simply stop developing AI?

THIS ANSWER IS DEPRECIATED, GO TOWhy don't we just not build AGI if it's so dangerous?

We could, but it seems unlikely. Each advance in capabilities which brings us closer to an intelligence explosion also brings in profits for whoever develops it (e.g. smarter digital personal assistants, more ability to automate cognitive tasks, better recommendation algorithms for social media). The incentives are all wrong. Each individual actor (nation or corporation) fears that if they stop, they’ll just get overtaken by others who are more reckless.