A new book, The Age of AI and Our Human future has just been released. The authors are Eric Schmidt, Daniel Huttenlocher and, astonishingly, Henry Kissenger. Now 98, Kissenger keeps popping up everywhere, as if to prove that the longevity of the human brain can seem unlimited, at least in some people.
I recently listened to an interview with Kissenger by The Economist magazine. If anything, he seems somewhat more coherent than he used to be. I was never a fan of his disastrous Vietnam policies when he was Secretary of State for Presidents Nixon and Ford. However, his breakthrough in the opening of China to the West at that time was where he made his mark, and it has kept him in demand as a China expert to this day.
In a review of the book by Kevin Roose in the New York Times on December 12, 2021, Roose gives a less than enthusiastic review. He had been hoping for more from a book about artificial intelligence by a 98-year-old diplomat, a former Google chief executive, and an M.I.T. professor, an intriguing coalition if there ever was one.
Tongue in cheek, at least somewhat, Roose has the brainstorm idea of having an A.I. writing program help with the review. Using Sudowrite (there's a great name), the app uses GPT-3, a cutting edge A.I. system. Provide a snippet of text, and GPT-3 will try to complete it, using what it has learned from billions of examples of other people's writing, along with the help of a supercomputer with 285,000 processors, and a neural network that ranks among the world's most powerful A.I. engines.
As a writer myself, I have to wonder what I wasted my life on for the past forty years when I could have simply logged onto GPT-3 instead. Roose provides us several paragraphs of the GPT-3 review, and it frankly reads quite reasonably. It's doubtful any reader could have told the difference...at least after the system's first somewhat clunky efforts: "The book which you are reading at the moment is a book on a nook, which is a book on a book, which is a book on a book on a subject, which is a subject on a subject, which is a subject on a subject." "GPT-3," writes Roose, "had gotten stuck in some kind of odd, reclusive loop."
But apparently the A.I. program simply needed to get warmed up. Just like humans, evidently, a few cups of coffee and stretching exercises were needed. The program soon began producing cogent analysis. So don't get your hopes up, fellow writers. Our A.I. replacements are eagerly waiting in the wings.
What The Age of AI seems to be telling us is that while A.I. systems can be clunky and erratic, they are improving quickly and will soon surpass humans in a number of different areas. At that point, the authors tell us, A.I. will "transform all realms of human experience."
I found the participation of Henry Kissenger in this endeavor somewhat surprising. He has little in-depth knowledge of this new technology. However, his own experience on the stage of world politics suggests one futuristic need of the human race. Perhaps we could turn over the running of world government to A.I.
It could hardly do a worse job.