Can we really put the brakes on AI development?

Some tech execs want a ‘pause’; the US president wants voluntary vetting – but can anything help keep AI under control?