In Review - I, Robot

Published 📅: ....
Last modified 📝: ....
Location 📍: Boston, MA

Share this post on BlueskySee discussion on Bluesky

I recently finished reading I, Robot, I think I had read small excerpts from it in some english literature class in high school at some point, or maybe I've read the book before and just didn't remember it. I don't think any other book that I've read comes even close to how this book resonates for me - and yet I don't think it's my all time favorite book, nor would it end up in my top 10, -20, or even -100 books that I've read.

With that being said though, I do think that everyone should give this book a read, not just because it's pretty short (right around 200 pages) and a really light read, but also because its almost prescient in how it talks about robots like they are the current LLMs and current day AI.

Three small anecdotes that stand out to me;

1. My reaction while reading this book
2. The "Liar!" chapter/short story
3. "The Evitable Conflict" chapter/short story

My Reaction

First up, my reaction to reading this book. I'll be brutally honest - I picked this book up hoping that it'd be effectively the same storyline as the 2004 summer block-buster I, Robot movie with Will Smith. Sorry to burst the bubble for folks that maybe started reading this blog post thinking it'd be the same and then went out and impulse purchased it from Amazon your local bookstore, but this is not the same storyline as the movie.

The movie definitely connects with some aspects from the book, sharing one of the main characters, and pulling from a few plot points along the way.

Honestly - after about 10 pages I thought I was going to put down the book and move on to the next, and it wasn't until the next chapter and then finally the last chapter in the book that completely reversed my perspective on this book!

The "Liar!" Chapter / Short Story

This chapter immediately reminded me of the ChatGPT 4o release, wherein many folks realized that the new model release was continuously glazing (maybe a risky click for some) you as you interacted with the model. That issue became so popular that OpenAI had to put out a blog post talking about how they were handling the issues and moving forward from there.

This chapter was almost a one-for-one prediction of what happened with the 4o release, and really highlighted one of the core concerns many have raised with increased use of LLMs by so many people.

"The Evitable Conflict"

One of the best parts of this finale to the book - is that it really leaves it up to the reader to extrapolate with where things are heading (in the book, but now also in real life).

I feel like this represents one of, if not the most central, AI Doomerism perspective, that AI / AGI will progress to the point where it is no longer being guided by humanity but instead starts to make decisions to align humanity to it's perspective future.

I really hope that AGI doesn't pick up our line of work where we're at now with AI alignment and continues it towards "human alignment" - but only time will tell.

Tags:

Related Posts