The potential is huge. That's why the beancounters are so hot on this one. Anything that structurally lowers operating cost by that much is a no-brainer
At what risk though?
Like magic a whole bunch of problems that never existed will start appearing. Changing the whole infrastructure and foundation of aviation doesn't really do alot other than solve one problem which is human error and replace it with a litany of new ones which have traditionally be mitigated by humans operators.