Thanks for the link, although I can’t find a list of books – I see a couple in there (primarily Asimov), but it’s mostly links to magazine articles and interviews.
I agree; I’ve read a lot on that topic.
It’s like they are deliberately hiding the gems, isn’t it? You can find Hogan’s book already mentioned by Harmonious Discord there (recommended) and Algis Budrys’ Michaelmas, which must have been exceptionally modern in its time .. but not anymore.
athelas recommended Permutation City – and I’d add Egan’s Diaspora; the AIs are at least interesting varieties, insofar as they are the result of human copies.
If you don’t mind to read a book that’s meant for younger readers, don’t miss Genesis by Bernard Beckett (but it’s much better to listen to the audio version). You will realize pretty soon what it’s really about but it doesn’t take away from a pretty good discussion of the question what it means to be human.
Alastair Reynolds’ Revelation Space Trilogy is hard science fiction that deals to some degree with AIs and the “artificialization” of human beings.
Most books are simply incredibly naive when it comes to AI - but you already know that.
I’d like to mention Exegesis by Astro Teller.
It’s about a program (EDGAR; Eager Data Gathering And Retrieval) that is supposed to skim the internet for info. Every once in a while, it sends an email to it’s creator, Alice. For a very long time, all it’s input resembles something like “the middle east is located at www.uruk.gov” however, one day she receives an email that simply reads “Hello, Alice”.
The book is written as a series of emails from Edgar to Alice and back. Its speech is stilted at first, and realistically improves. As Edgar starts out as a text interpretation tool, it’s mystified by images, and only has a very limited understanding of the real world. The characters also discuss if Edgar is in fact an intelligent being, or a program really good at simulating. The moral and philosophical consequences are similarly adressed.
I think it’s a very, very good book.
Missed the edit window:
I forgot to add Rudy Rucker’s Ware Tetralogy - not every books is worth a read but they are definitely not naive in their presentation of AI.
Oh yes, Exegesis is a good one, MostlyClueless.
This doesn’t sound like a likely scenario. Long before we are emulating a human brain, we will be simulating its various functions. We already do that, and there are cases where a computer ‘out thinks’ humans now. The leap into real AI will come when computers can self improve themselves continuously. Even at that stage, they are unlikely to function the way that human brains do, and that first human simulation will be created by computers. Even though I think we will one day be able to emulate a human brain, I would bet on AI reaching the equivalent of human brain function using techniques less susceptible to schizophrenia.
The robot concept makes for good stories, but seems unnecessary. AI doesn’t need a body hands and feet, and robots don’t need to be controlled by their own brain. All of the AI systems may become one large brain anyway, and eschew human concepts of individuality.
Asimov’s stories looked at the three laws a lot, but they also were about the integration of robots into human society. 2, 3, 4, and 5 from your example were integral to many of the stories.
TriPolar – you seem to be giving a WAG as fact. What you say is certainly plausible. So is the example I gave. In fact your comments about simulating various brain functions separately (which is not at all the type of AI I am interested in, but when combined could become it instantaneously) as well as your mentioning the non-linearity of the whole process seems to strongly support the type of scenario I am describing, at least in spirit. In any case this discussion is completely irrelevant to my OP, where I gave but an example of what in bare-bones form might constitute what I was interested in and consider realistic enough.
Also, please, so that we don’t derail this discussion into an argument about Asimov – I have already read the whole series recently and there is no point in you pointing it out to me. Full stop. I did not find the series meaty, to say the least. I had looked forward to reading it for a long time, but was really let down. It really doesn’t even come close to what I am interested in and have described in the OP. At the very least he doesn’t create one history or coherent narrative – the stories are isolated, partial and often unrealistic discussions of some of the bullet points I gave, but mostly focusing on cute aspects of the 3 laws somewhat narrowly, and consists primarily of childrens detective stories, primarily in the far far future.
Hi all,
I’ve checked out all the recommended books and added a bunch to my wish list. Some look really good! Since none really fit the description of my OP (except perhaps The Ware Tetralogy, as far as I can tell from wikipedia), let me try to be a little more clear.
I think I’m interested in what could be described as a “realistic” “alternate-history” style narrative. And by “realistic”, I mostly mean “the way real people would behave and react.” I haven’t read the recommended Exegesis, but even though it’s not a “history” (it sounds more like an isolated story about the birth of an AI), it does sounds interesting in that I think “what would I do in that situation.” As long as this Alice Wu is not depicted as a moron the way most movie protagonists are, this sounds like a fascinating read.
You could try Mindscan and The Terminal Experiment by Robert J. Sawyer (about the ability to copy human brains, not create fresh ones), and I second the recommendations for Permutation City and Diaspora by Greg Egan. And probably Egan’s new book Zendegi as well, though I haven’t started reading it yet, but it deals with a fictional “Human Connectome Project,” about a massive Human Genome Project-style attempt to map the connections between every neuron in a human brain.
ETA: Oh, and of course Robert J. Sawyer’s latest trilogy (due to be completed this spring), WWW, deals in great detail with the spontaneous emergence of a conscious entity on the Internet, and how humanity reacts.
Ok, I didn’t mean to start a debate on the subject, I was just trying to delineate what you are looking for. I agree with much of what you say about the Asimov stories. And I see you consider a continuing storyline centered on AI as important. I’ve been off the SF for a while now, but I’ll ask a few of the continuing users I know for any recommendations, and maybe I’ll recall something I’ve read.
Are you looking for the entertainment, or an author’s concept of how the future will play out. It’s always difficult with SF because some readers are more interested in either the S or the F.
Thanks TriPolar. I’m mostly just looking for entertainment, but I only really find very hard sci-fi entertaining (I generally can’t stand fantasy – seems too easy to fill plot holes etc when there aren’t any hard rules). And though I’m a real science-geek by trade, I’ve mostly stayed away from sci-fi my adult life because a lot of the stuff I have been exposed to is a little sophomoric (though I don’t mean to belittle the genre – I have read some really fantastic sci-fi books, but it is really hit-and-miss for me). I mostly read non-fiction or history, so that is probably where I am coming from. I’ve always yearned for a book like I described in the OP, read perhaps the way you might read a good history book, and figured it might be common-hat to a bunch of sci-fi readers.
When you find it, let me know. I think you could be pleased by Rudy Rucker’s work insofar as he is able to combine a comprehensive understanding of the AI research with an imaginative extrapolation. But the result is not always a coherent story. Or a well written one – though usually better than the average in Science Fiction.
I remembered another book that you might find interesting: Accelerando by Charles Stross. It deals with the idea of technological singularity that was made popular in futurology by Vernor Vinge and Ray Kurzweil.
This topic describes a period that is technologically too far advanced to fit your bill, but the stories within the book encompass the whole process by dealing with three generations of a family involved in the events.
Yeah, I accidentally wandered into The Singularity is Near in a B&N when it first came out. I sat down on one of their comfy chairs thinking I’d skim it, but I ended up reading the entire thing right there! Good stuff.
Check out:
When HARLIE Was One, by David Gerrold (1972).
Society of the Mind, by Eric L. Harry (1996).
“Oracle” (novella) by Greg Egan (2000).
All deal with the theme of AIs discovering and pursuing their own agendas.
Whoa there, I said I liked hard sci-fi, but damn! 
Just pretend that part is Star Trek technobabble – it might as well be, for plot purposes – and read past it.
Can’t believe I’m the first to mention 2001: A Space Odyssey. The story, of course, is one of the worst-constructed classics in SF: We’ve got this cosmic plot about superior ET intelligences using these obelisks to accelerate and guide human evolution . . . and sandwiched in the middle of it, on the journey from Earth to Jupiter, this completely unrelated plot about an AI, HAL-9000, malfunctioning with deadly results, which could have been made a separate novel/film, or could have been left out entirely without changing the ending much; its only function, WRT the ending, seems to be to allow only one astronaut to reach Jupiter alive, and have to deal with the strangeness there all alone.
But the book, which I won’t spoil here, makes it much clearer than the film just why HAL malfunctioned (hint: it wasn’t quite entirely his fault, nor was it any kind of mechanical failure or damage; so it should be of some interest to the OP).
[hijack]
Don’t tell me you have never read about the connection between HAL and the ever present creation myths in 2001?
Apart from HALs association with the stories of Prometheus, Frankenstein and Nietzsche’s Thus Spoke Zarathustra, we have also the multiple connections towards Homer’s Odyssey. Bowman was often described as a modern Odysseus (the name Bowman is almost too obvious, isn’t it?) and HAL as the cyclops.
.. hm, come to think of it, not a hijack, not at all.
John Varley’s short story “Press Enter” is not exactly what the OP is going for, but touches on many of those issues, and is a chillingly good tale to boot.
Charles Stross’s novel Saturn’s Children is about Three Laws-governed androids forming a new society throughout the Solar System after humanity suffers some unexplained extinction event; some interesting concepts, although the storytelling is subpar, alas.
Larry Niven story about building an AI computer. From Draco Tavern series of short stories.
THE SCHUMANN COMPUTER
Funny little parable about the price of too much knowledge, as the world’s most powerful AI figures everything out…and it ain’t 42. Could be seen as a satricial rebuke to the idea of divine omniscience, as well, though that may not have been what Niven was after. This was the first Draco Tavern story to see print, back in 1979, contemporaneously with Hitchhiker’s Guide; I’m kinda guessing the two probably didn’t influence each other! Also appeared in Convergent Series
Wow that’s a super specific plot. Not sure about that. The Matrix animated DVD has a bunch of interesting background on the history of the Matrix world with a lot of what you are talking about, you might check that out, though it’s not a book.
The most realistic book I’ve come across about AI is The Lifecycle of Software Objects by Ted Chiang, which is based on the idea that real AI would have to be basically reared and trained.