Back Cover Blurb from Klara and the Sun by Kazau Ishiguru:
Here is the story of Klara, an Artificial Friend with outstanding observational qualities, who, from her place in the store, watches carefully the behavior of those who come in to browse, and of those who pass on the street outside. She remains hopeful that a customer will soon choose her. Klara and the Sun is a thrilling book that offers a look at our changing world through the eyes of an unforgettable narrator, and one that explores the fundamental question: What does it mean to love?
“As I say, these were helpful lessons for me. Not only had I learnt that changes were a part of Josie, and that I should be ready to accommodate them, I’d begun to understand also, that this wasn’t a trait peculiar just to Josie, that people often felt the need to prepare a side of themselves to display to passersby – as they might in a store window, and that such display needn’t be taken so seriously once the moment had passed.”
― Kazuo Ishiguro, Klara and the Sun
from Goodreads
I was so excited about this book. Perhaps that’s where things went wrong.
I found the title by way of someone on social media that I typically enjoy on her list of favorite books of the year. I was pleasantly surprised to find that it was a story with artificial intelligence as a central theme written by a Nobel prize winner. I have been reading about artificial intelligence, both as non-fiction and fiction as a theme over the past year.
This is a blog post exploring what did not work for me about this book.
I enjoyed the beginning of this book. I liked the blank slate of the totally new, fresh out of the box Artificial Friend (AF), Klara.
At some point, I had to check what age this book was written for, because Klara continues to act so child-like and naive that I though it might be a young adult book. To me, it actually has more potential as a YA book. It could be a tale for kids coming of age as AI does, about how to think about themselves, humanity, and artificial intelligence. But it did not go in that direction at all.
There is a weird and disturbing subplot about the mom asking Klara to mimic Josie’s disabilities so that the mom can have a copy of her if she dies. This could be interesting, if it were written differently, but it is not to me. In another execution it could be an inquiry into what it means to be a shadow or an echo of a human you love, rather than simple mimicry; or a better window into how humans think of each other and handle, or don’t handle, loss. I found the current execution narcissistic, weird, and cruel. How dare my sick child inconvenience me on my day off? I will just take this robot when I want to go to some waterfall that means something to both of us! If this is what it means to love, I am depressed. Josie is going to need some serious therapy.
Then, there is a weird mix on the technology side between including references to it, like Klara describing her vision in boxes, and not considering it at all, like no mention or though of feeding back into or advancing the other AI algorithms based on Klara’s learning. So many human hours of coding, annotating speech, image processing, robotics for walking, have gone into bringing us to the precipice of benefitting from AF. We are going to live with an instance for years and not feed any of that learning or data back into the model? Now that sounds wasteful and naive.
So it makes sense to me, or I interpret the book as, Klara is solar powered and therefore obsessed with her power source. That sort of insight is kind of fun. What will AI “worry” about when their worries are separated from those of humans? I originally guessed they would be hungry for data, but power makes sense too. Power always makes sense as a thing sentient beings might want. Then Klara, the AF, starts praying to the Sun.
This, to me, is the opposite of what a good book about what a book about how artificial intelligence plays out in our live might look like. That the Sun would be Klara’s higher power makes sense. I understand that AI is our creation and picks up our biases. Yet, the idea that an AI would pray, much less on behalf of a human, was the final detail that did not work for me. It might be a personal Uncanny Valley or the wrong personification of a machine for me.
I read the Wikipedia plot summary, and I liked where things were headed even less than what I had read so far, so I quit reading. I was so disappointed on a number of levels.