I recently attend a conference of AI and Sustainability and came away with a different perspective. I have been doing this investigating for some time at other ‘AI‘ discussions and have resolved an understanding that what is being described as AI, as in Artificial intelligence, is in reality, only an example of the second stage of industrialist Revolution, the electronic analysis of (big) data to control manual (automated) machines.
The only element of the current AI (buzz) business that could be considered intelligent, is the intake, and analysis of data as a front end user interface to existing machine technology. Fundamentally the ‘Intelligence’ is only a programmed response to data changes which adapts limited machines, to adjusting to changes in the data, that it is designed to digest.
There is real AI, real sentience research going on, but it is years away from real applications. What the public has been seeing in Alexa, SIRI, Google Assistant, MyCroft and Open Assistant is just programmed ‘Simulated Intelligent‘ (SI) not real cognitive intelligence.
In the more extensive examples of ‘AI‘ or SI the ‘intelligence’ is just more complex, and adapts to more complex data sets than a simpler one. A device that converses with you is doing more analysis, of voice patterns, language syntax, and context. It does NOT understand what you are saying, it operates only within the limited, but expandable, programmed responses. So if a self driving car kills someone, it’s not the fault of the AI/SI it is the programmers fault for bad programming.
This whole interpretation of SI rather than AI will make for a great deal of Legal entanglements that will last for years. And meanwhile the development of real AI will develop to the point that when (I believe in when) there is a real AI, a sentient, cognitive Intelligence we will have to revisit the entire meaning of being alive.
And maybe then, we can worry about people losing their jobs due to AI.