Adopting AI: The People-first Approach

"Few books so artfully blend the optimism of Kurzweil with the caution of Bostrom with the pragmatism of Mollick." (Linus Caldwell, Alphaverse Capital)In classical Greek mythology, Prometheus stole fire from the gods and gave it to humanity before being damned for all eternity. The moral of the story: the gift of scientific progress and technology is a double-edged sword.If the hype is to be believed, humanity is on the cusp of a technological breakthrough of Promethean impact: Artificial Intelligence. Google’s Sundar Pichai claims that the impact of AI will be “more profound than fire and electricity.”Critics, however, are divided on the prospects. Will AI bring about a techno-capitalist utopia where scarcity is eradicated, disease and aging are conquered, and the deepest mysteries of the universe finally revealed? Or will it create a dystopia of the sort familiar from countless science fiction classics: The Matrix, Terminator, Dune, or 2001: A Space Odyssey?To illustrate the knife-edge on which humanity finds itself, Adopting AI starts with seven scenarios, each of which has a catastrophic and a utopian outcome. The authors invite the reader, at the beginning and end of the book, to assess where they think humanity is, and in which direction it might be heading.In the authors’ view, the answers aren’t predetermined by technology but depend on the choices individuals, organizations, and societies make in the coming years. Paradoxically, the more powerful the technology, the more important human reason and moral judgment become.Gibbons and Healy argue that successfully and safely adopting AI means recognizing just how different this transformational technology is from the traditional computer systems with which we’re all familiar. In their provocative take, AI is best thought of not as a technology at all, but as an intelligence: the smartest, dumbest, strangest colleague you’ve ever had.The principles forPeople-first AIadoption flow from the fundamental nature of AI: unpredictable, opaque, and prone to bias. Yet