Published on December 31, 2024 10:01 PM GMT
Book review: Genesis: Artificial Intelligence, Hope, and the HumanSpirit, by Henry A. Kissinger, Eric Schmidt, and Craig Mundie.
Genesis lends a bit of authority to concerns about AI.
It is a frustrating book. It took more effort for me read than it shouldhave taken. The difficulty stems not from complex subject matter(although the topics are complex), but from a peculiarly alien writingstyle that transcends mere linguistic differences - though Kissinger'sGerman intellectual heritage may play a role.
The book's opening meanders through historical vignettes whoserelevance remains opaque, testing my patience before finally addressingAI.
Risks
When the book gets around to discussing how AI will affect our future,it's mostly correct about AI being a big deal, with occasionallyappropriate hints about why there are big risks. But it's frustratinglyabstract and vague. Some examples:
we might become extinct.
Would networking intelligences make their processes more opaque thanthe processes of lone intelligence? ... would we be able to assessthem on a spectrum of good to evil? Or would they operate on aninformational basis - extracted at superhuman speed ... - that wouldconfound our ability to judge their behavior? Would that lead usfurther into a cycle of passivity?
Today, in the years, months, weeks, and days leading up to the arrivalof the first superintelligence, a security dilemma of existentialnature awaits.
I see hints in that quote that they think the threshold ofsuperintelligence will be well enough defined that it can be attributedto a specific day. I find that suspicious.
Genesis compares our preparedness for AI to the preparedness of Aztecsfor the arrival of conquistadors.
One area where the book briefly feels clear and novel is when itdiscusses the future of war, notably observing that humans may becomeless targeted simply because they'll be irrelevant to militaryoutcomes.
The book provides only weak hints as to what considerations areimportant. It often feels like there's a missingmood -e.g. it's hard to tell whether the authors think human extinction wouldbe a bigger deal than the end of democracy.
Present Day AI
The weakest parts of the book attempt to describe current AI. Too manyof those claims look like claims that were discredited several yearsago. It was published a year after Kissinger's death, so likely some ofthe problem is a long delay between when he wrote those parts andpublication.
But there will be phases in the evolution of AI when mechanicalintelligence may feel eerily similar to the intelligence of theanimals.
I'd say that "prediction" was plausibly true of the best AIs for abrief time around 2021 or 2022. Now AIs seem more like human children.
Lately, AI researchers have devoted serious attention to the projectof giving machines "groundedness" - a reliable relationship betweenthe machine's representations and reality
This was true in 2022, but it has been increasingly treated as a solvedproblem since then.
Other Thoughts
Will we become more like them, or will they become more like us? ...Answering it remains our first and most necessary task.
The authors express cautious optimism about brain-computer interfacesfacilitating human-AI symbiosis. That suggests either an overestimationof neural interface potential or an underestimation of AI's rapidadvancement.
Under this definition, can AI itself possess dignity? Likely not - forAIs are not born, do not die, feel neither insecurity nor fear, and donot have natural inclinations or individuality such that conceptionsof evil or good could be considered "theirs". ... they should betreated, philosophically, like literary characters.
This feels like a confused mix of half-assed morality and limitedunderstanding of where AI is headed.
Genesis refers to Nick Bostrom and Eliezer Yudkowsky without criticizingthem. Combined with Kissinger's reputation, that will cause somepolitical and military leaders to take the risks of AI more seriously.That makes the book somewhat important.
People should read this book if they respect Kissinger's forecasts muchmore than they respect the forecasts of people connected with techcompanies.
Discuss
