AI depends on mathematics, algorithms, computers. Knowledge creation depends on making conjectures and refutations and serendipity. Human intelligence is neither ultimate nor absolute. Homo sapiens are not the ultimate manifestation of the genus Homo. Homo sapiens are still evolving and are likely to speciate into a more powerful life form in the future. The best of human knowledge is captured in human designed axiomatic systems. These systems are not perfect as we learn from Gödel Incompleteness Theorem, Turing’s halting theorem and Chaitin’s Algorithmic Information Theory. All future AI systems will need to account for these and many more yet undiscovered aspects of human intelligence derived knowledge. These aspects are not touched upon Thomas Davenport’s The AI Advantage. This is not a criticism of the book because Davenport has sought to address a very different, down-to-earth audience of business entrepreneurs who are unlikely to possess deep knowledge of the functioning of the human brain, its intellectual abilities (and disabilities), its innovative ability to create artificial intelligence, etc. What AI does today, even half-century ago would have appeared superhuman and miraculous. Then, intelligent activity was assumed to be unmechanizable and therefore not a threat to ‘intelligent’ humans. It all changed when IBM’s Watson beat human experts in the TV game Jeopardy in 2011 and Google’s AlphaGo beat world champions in the game Go in 2017.
User
Font Size
Information