Microsoft’s AI needs just 3 seconds of speech to bring people back from the dead

Microsoft has announced a new tool that it says is capable of replicating a person’s voice after analysing just three seconds of sample audio.