It has become fashionable in the last year or two to say that Artificial Intelligence (AI) has finally come of age, implying that some big software breakthrough had been made so that it has now become the final putsch in the great take-over of people’s jobs by computer. Not so. Automation is still proceeding at the steady pace it has done since the first cotton spinning machines of 1780.
Instead, the recent excitement about AI is the realisation that a whole new class of problems can be solved involving finding discrepancies — or matches — within truly vast amounts of data. For example, driverless cars are not driven by any sort of upgraded programming but by being able to access billions of visual memories of the expected roadside environment and to compare them with what the car is actually seeing around it as it drives along.
If AI researchers are not yet able to simulate human processes of cognition and associated self-awareness, they’re working on it! They also have a term for it to distinguish it from ‘ordinary’ AI. It is Artificial General Intelligence (AGI). It is AGI, not AI that will be the real killer of jobs — those with unspecified as well as routine elements in them. We recognize that a great many more jobs are going to be swept away by AI in the next two decades, but the fear is — will the rest also go by way of AGI?
But will AGI actually come about? I have the same reservations about it as when it was merely called AI.