For those of you that like Sapiens, Homo Deus by the same author, Yuval Noah Harrari, picks up where Sapiens left off. Sapiens was a backward look, and Homo Deus is a future cast of where we may end up given the predominant frames that we have developed as humans.
Giving a tour de force of historical review, Harrari goes back in time to explore how humans have gone from belief in multiple gods (animism), to singular god (theism), to believing humans are god (humanism). His starting point for the futurecast is that we as homo sapiens have embraced a new religion of humanism – that in which we are the ultimate creator and working to achieve ends and means that benefit humans in the current life.
Harrari proposes that humans have “beaten” the three traditional drivers of death and destruction: war, famine, and plague. We now are focused under the religiosity of humanism, on finding ways to engineer more pleasure, create more perfect humans and achieve divinity. In the attempts to do so, he proposes that humans are working towards AI and singularity (the blend of man and machine) that will ultimately turn humans into ever more powerful algorithms to the point that humanity will lose its ability to be relevant and meaningful in a world run by machines and equations. This could result in the creation of a significant split in the humans that have value and worth – the new elites – as highly integrated and sophisticated machines that control much, and the rest of us. As this transition and move becomes intensified, the value of data will reign supreme and our focus will shift from humanism to dataism.
It is an interesting read but I found myself wandering. While I agree with Harrari that the march towards AI is fraught with challenges and pitfalls, I find his writing style unapproachable. It comes across as he is certain he is the smartest person in the room and so I felt a distance to his arguments. I agree with the premise – that we must think very deeply about whether we should head down the paths of greater AI development and singularity. I also found myself wishing he had summarized his points with less detail. This is one of those books that could have been written in half the amount of text.
Verdict: If you stay up at night worried about the future of AI and humanity, give it a read, otherwise take a pass.