None
4 stars
Very minor, non-story spoilers in this review.
I don't know exactly when or how this book came to be on my to-read shelf. It's an ex-library copy, from a local library, so it's more likely I bought it impulsively, as opposed to online after seeing it on a list of VR books or something. The book feels extremely relevant to my interests, as it deals heavily with the topics of VR technology and AI ethics.
The story takes a while to get into the VR. Part 2 starts around page 95, and at that point the story jumps forward in time from 2012 to 2027. In the first timeframe, 2012, one of the characters is working on mapping the brain of a small bird. The lab she works in painstakingly digitizes several bird-brains, and she works to create a digital model that is a hybrid of the scans. In 2027, …
Very minor, non-story spoilers in this review.
I don't know exactly when or how this book came to be on my to-read shelf. It's an ex-library copy, from a local library, so it's more likely I bought it impulsively, as opposed to online after seeing it on a list of VR books or something. The book feels extremely relevant to my interests, as it deals heavily with the topics of VR technology and AI ethics.
The story takes a while to get into the VR. Part 2 starts around page 95, and at that point the story jumps forward in time from 2012 to 2027. In the first timeframe, 2012, one of the characters is working on mapping the brain of a small bird. The lab she works in painstakingly digitizes several bird-brains, and she works to create a digital model that is a hybrid of the scans. In 2027, she's had several career jumps, but sees that there are now human brain-scans of equivalent granularity, and she realizes it's only a matter of time before someone works to use her previous technique on this new scans. Having done it once previously, with the birds, she is able to be the first with these new human scans, and uses the resulting models at the company where she works, a sort of game system company that sells hardware VR platforms called "castles" that are fully immersive.
The book was published in 2010. This was only a couple of years before the first Oculus kickstarter in 2012, although of course plenty of VR headsets had existed for decades previous to that – they were just not consumer-focused.
So the book does a pretty good job imagining VR... although it doesn't imagine it in people's homes. More as space to rent in specialized electronics shops. VR arcades are a thing, so this is not terribly far off the mark. Additionally, who is to say things will not shake out to go more in that direction in the next 4 years.
But where the book completely misses the mark is with its imagining of AI. LLMs have completely changed the entire AI landscape seemingly overnight. I do not fault Greg Egan for not realizing they were coming. Most of the uses for these brain-scan-based AI in the book are pretty much entirely do-able with LLMs now. (Although LLMs can emulate a person, they will not have the personality or memories of any person, and that enhancement feels more than 4 years away to me at this point.)
In the book, there are hundreds of protesters every day outside the office of the VR company for exploiting AI based on human brains. When we do develop the technology to do this, I seriously doubt we'll see any reaction along similar lines. It'll just be business as usual.
