Pikkukuvaa napsauttamalla pääset Google Booksiin.
Parallel Realities: A Turing Fiction
Tekijä: K.R. Simms
Ei tämänhetkisiä Keskustelu-viestiketjuja tästä kirjasta.
The story itself was rather interesting. The two realities presented throughout the book were compelling as futures that had deeply integrated technology into all aspects of life. I do feel that the way in which the details of this world were presented could have been better. Namely, we often just have a character just describe to Dawson what some technology is, rather than have him learn through experiences, which felt awkward.
There were several things throughout the book that did not make sense in the moment. This included, but was not limited to, details about machine learning (and CS in general) that was off. Much of it ended up being explainable after reading the last chapter, but before that I found it rather frustrating & distracting. However, I imagine that for a reader that does not have a background in machine learning/CS, the oddities with ML would not be an issue.
My biggest complaint about this book is the writing style. ML-generated language is generally clunky, repetitive, & occasionally nonsensical. While this book (supposedly) has both human-written and machine-generated text, I found this problem with language was prevalent throughout. One could argue that the author was trying to match the machine's writing style in an attempt to throw off the reader on which parts were written by which contributor. This doesn't change the fact that it made for a worse reading experience. Yes, getting good language from a ML model is insanely difficult, but perhaps the author could have worked more with whatever model he was using to get better output.
Generally, this is a decent plot & the idea overall is interesting, but the execution did not really work for me. However, I'm coming at this with a decent CS/ML background - a reader with less background might be able to enjoy this more
Simms includes sections wholly written by an AI, “a conversational AI system that listens, learns, and challenges,” or I have to assume he did. The narration is somewhat choppy and clunky throughout, so I wasn’t sure which segments were human-written first drafts or AI-generated. And, unfortunately, I was sometimes distracted by contradictions in the narration.
But aside from the initial challenge, the novel eventually provided my rational examination with many intriguing potentials in AI-human interactions: in government, law, and of course human relations. Trust is important in human relations, and that’s precisely what the main character doesn’t have -- with an AI-implant in his brain to save him from a catastrophic brain injury, he doesn’t know whether his “experiences” with his father and his doctor, among others, are “real” or not.
So for me Parallel Realities is about someone becoming extremely paranoid because he can’t determine real from manufactured experiences, read by a person already somewhat paranoid about AI who can’t determine the human from the manufactured prose.
I received an early review copy from Librarything. This is an honest review.
Näyttää 1-5 (yhteensä 17) (seuraava | näytä kaikki)
Kirjastojen kuvailuja ei löytynyt.
LibraryThing Early Reviewers Alum
K.R. Simms's book Parallel Realities: A Turing Fiction was available from LibraryThing Early Reviewers.
Amazon Kindle (0 painosta)
Audible (0 painosta)
CD Audiobook (0 painosta)
Project Gutenberg (0 painosta)
Google Books — Ladataan...
Beyond these style issues, my only real complaint about the story is the lack of a clear ending. Perhaps I am too literal and don't feel satisfied unless the nature of the reality of the story is spelled out. I never liked Zen koans either, but just when I felt like we were getting somewhere in the plot, the book ended. ( )