I suppose the AI debate is similar to that in the 40/50’s concerning Nuclear Fusion – Good or Bad. The trouble with AI is that whilst we may THINK we control the ON/OFF switch, which we clearly do at this stage, how long will it take these machines to self-learn to protect there own functionality. If they ever reach that point then the sci-fi movies may prove to have been correct.
Used wisely AI may accelerate medical, scientific, defence, engineering and many other areas where knowledge is the foundation for progress. However, there needs NOW to be a Universal Control mechanism to ensure AI is constrained for the good of mankind. For example, who needs ‘driverless’ cars?
Watch as this area of technology may just be the best or worst thing we have so far developed.