The main dish this time is Greg Egan’s novel Zendegi, a rich brew of near-future Iran, metaverse gaming, AI-modeling, mind-uploading and family tragedy – very human. It’s a bit of an atypical title in Egan’s oeuvre, and totally different from 2008’s Incandescence.
I’ll end with an appraisal of Dispersion, a fairly recent 158-page novella about a breakdown in a pastoral-ish society with 6 factions that operate more or less in different dimensions, out of sync most of the time. Egan demonstrates that the scientific mindset is the way out, not distrust and tribalism.
I enjoyed Zendegi, even though the novel could have been better. Egan offers a story that tries to do a lot, which makes for a diverse reading experience. At first it is a near-future political thriller set in Iran, and it morphs into a story that combines a family tragedy with stuff about differing cultures, AI and mind-uploading.
Egan admits in his notes that the first part of the book “was always destined to be overtaken by reality”. He finished it “in July 2009, a month after the widely disputed re-election of President Mahmoud Ahmadinejad”, followed by massive demonstrations and brutal crackdowns. Even though what Egan described in a fictional 2012 didn’t come to pass, he expressed the hope “that this part of the story captures something of the spirit of the times and the courage and ingenuity of the Iranian people.” It is no spoiler Egan’s future Iran more or less embraces modernity again.
Martin Seymour is an Australian journalist covering Iranian elections. The 2012 part is one third of the novel, and basically serves as a set-up for the second part. That is set in 2027-2028, and it turns out Martin remained in the country, married an Iranian woman, fathered a son. This information is on the backcover as well, but it interfered with reading the first part, as I wanted to get to 2027 sooner than if I hadn’t known what was in store.
The book’s other main character is Nasim Golestani, an exiled Iranian scientist who returns to her homeland to help rebuild it. She is part of research to construct a detailed map of the human brain, and gets into game developing later. The title, Zendegi, refers to an online VR game that Martin’s son plays.
Egan touches on themes also explored by Neal Stephenson’s 2019 Fall, or Dodge in Hell and Richard Powers’ Bewilderment. He is at least as successful. The parts that explore the technicalities of constructing a software human brain are really interesting and well-done. At the same time, especially in the 2027 part, it is mostly handwavium.
As for the family tragedy, Egan had me crying twice, so that amounts to success – even though I feel he could have explored these issues even deeper: especially the portrait of his wife could have been painted in much more detail. Then again, for a novel of 329 pages this is a remarkable rich, sturdy novel. Adding more to the personal side of things might have resulted in a bit of bloat, but maybe 50 or 70 pages extra would have hit the sweet spot without turning this into a door stopper.
Ultimately, Egan’s focus is the construction of AI and forms of mind-uploading. The good thing is that Egan and his characters don’t aim for full-on AI or full-on consciousness. The scientists in the book try to construct algorithms merely based on the mapping of neural networks of the human brain. I’m not sure what I think about this kind of approach. I’ve read a few books on brains, but I’m not an expert. Either way, to me it seems a stretch that a software model would be able to talk just based on the connections it charted during brain scans – even though I tend to agree with Alex Rosenberg that there is no content as such in our minds, but only connected neurons firing. Getting from a chart of connections to a working algorithm seems science fiction – indeed.
Either way, that doesn’t hinder Egan’s ability to ask interesting methodological questions and questions about artificial consciousness: is it ethical to create one if it isn’t autonomous?
But the ethical questions remain theoretical: can we even build a consciousness? And what is consciousness to begin with? It would have been interesting to read more about Egan’s take on these matters, and the book would have been richer – even though there is a fair amount of nuance already, much more than Blake Lemoine recently showed about LaMDA.
Egan was ahead of the mainstream curve. He uses the term cishumanist – the term ‘cisgender’ was coined in 1994, but only added to dictionaries starting 2015, and I’ve only seen widespread use for a couple of years. And like Stephenson in Fall, he also describes the epistemological problems of the internet: bubbles and fake news, but he does so 9 years before Neal, and also 2 years before the bubble phenomenon was even named as such, like C pointed out.
I would have trimmed bits of Martin and his son Javeed playing the Zendegi VR game. It just detracts from the much more interesting issues: science & tragedy. And, like in Fall, the story-line of the fantasy setting simply isn’t that interesting. Aside from the fact that we don’t get that much time to bond with the setting and the software characters, it might be additionally hard because of a formal matter: we know it’s not a real story in the book, in a way making it doubly fictional (fake), and we have to suspend our disbelief twice: once for Egan’s main narrative, and again for the nested narrative in the game. Maybe because of that it is less engaging?
Aside from the science, Egan shows himself a keen observer of humanity. Zendegi is also a book about bias, and Egan is, via Martin Seymour, an honest, soul searching commentator on these matters, with a sensibility for diversity and bigotry seeped in respect for what it means to be human, never pointing fingers, but not flinching away from judgement either. Even though I only mention it at the very end of this review, this bias isn’t some trivial matter: it is at the heart of the entire narrative.
Definitely recommended for Egan fans, and because it is very accessible without being superficial, it’s also recommended as an introduction to his work.
This novella – limited to 1000 paper copies by Subterranean Press – was okay, and that’s no mean feat in and of itself.
Why don’t I write I liked it? It was too *one-dimensional*: Egan takes one idea, and builds everything around it. 2019’s Perihelion Summer was a much, much richer novella, a five-star read even.
Here, I would have liked a bit more focus on the characters, even though I have to admit they weren’t completely out of focus. I also didn’t think the main speculative idea of the story was that interesting, even though Egan again rigorously thought things through.
What saved it for me was the solid prose, and the excellent economy of writing. The mystery of the world was sketched expertly, and Egan did manage to draw me in, enough so I wanted to finish it, even though it looses a bit of its magic throughout.
As for the message: distrust and tribalism indeed will not solve most societal challenges this day and age, but when you invent a specific speculative scientific problem as the problematic setting, it’s obvious only science will get you out – even though the characters don’t necessarily perceive their predicament as scientific. In that sense, Egan is a bit self-serving and his parable doesn’t really work as a lesson: it kinda begs the question. It would have been more interesting to start with a problem that isn’t so obviously scientific in nature to us readers – but that would have taken an entirely different set-up.
If you don’t want to pay $40 for the limited hardcover, you can get the eBook for $5.