“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”

“Co-Intelligence” is a concise, research-based and easy-to-digest book that talks about the current state of LLMs, how they work behind the scenes, what they can and can’t do and where we should and should not use them. 

The author, “Ethan Mollick” is a Wharton Professor who became famous in the Tech circles in 2024 when OpenAI cited his prompts in their research posts. If you work in Tech or Tech-adjacent fields, I highly recommend his substack where he regularly talks about AI and LLMs for folks who are AI power users but not necessarily AI technology experts. 

“This is the paradox of AI creativity. The same feature that makes LLMs unreliable and dangerous for factual work also makes them useful.”

The first half of the book might not be new information if you use LLMs heavily and/or follow AI development regularly. It is still an excellent summary though - covering everything from the concept of Transformers (the seminal 2017 paper from Google) to the human reinforcement training and current challenges in pre-training that results in bias. But it is the second half that is truly worthwhile. The author discusses social & economical changes that AI has brought about already and will likely bring in the future. 

“AI is the first wave of automation that broadly affects the highest-paid professional workers. Plus, AI adoption is happening much more quickly, and much more broadly, than previous waves of technology. And we are still unclear as to what the limits, and possibilities, of this new technology are, how quickly they will continue to grow, and how ahistorical and strange the effects might be.”

It is a superb and engaging read - that can and should be read regardless of whether one has a background in computer science or in AI/ML. The open ended discussions are thought-provoking. The book does not overtly laud AI and neither does it veer towards a doomsday scenario. Overall, a nice balanced and academic viewpoint. A solid 4/5.

Book Review of The Murderbot Diaries #4 - Exit Strategy

“I don’t know what I want. I said that at some point, I think. But it isn’t that, it’s that I don’t want anyone to tell me what I want, or to make decisions for me.”

I have to admit, the fourth installation in the Murderbot Series really did connect back to the previous books and tied up everything nicely. The plot is a bit predictable in this one but if you have stuck with the series so far, you are already fond of the sentient humanoid robot (who keeps getting sassier and sassier as the series progresses), so the predictability of the plot isn’t a dealbreaker. Overall, an enjoyable series and would recommend it to any SciFi aficionado. 4/5.

Book Review of The Murderbot Diaries #3 - Rogue Protocol

Well, the “Rogue Protocol” definitely sticks to the script like the previous two books in the series. Crisp story, fast pace & lots of action with our endearing sentient bot in the center of it all. What is slightly annoying now is that even though I am three books into the series and enjoying it so far, I don’t see a long encompassing theme per-say. This is more like a set of stories - strung together because of the same protagonist. Maybe something overarching will emerge in the next books, one can hope. Still, Rogue Protocol itself is perfectly enjoyable - 4/5.

Book Review of The Murderbot Diaries #2 - Artificial Condition

“Sometimes people do things to you that you can’t do anything about. You just have to survive it and go on.”

You know what's more fun than a fully sentient humanoid robot with deadly weapons? A one that also has a heart of gold. Even though this all sounds like a cliche, the second installment of the Murderbot Diaries series did not disappoint. Just one issue - it ended abruptly, and the world building seemed annoyingly incomplete at times. Still the book as a whole was a breezy fun. "Artificial Condition" - a strong 3.5/5 

Book Review of The Murderbot Diaries #1 - All Systems Red

"I don’t know what I want. But it isn’t that, it’s that I don’t want anyone to tell me what I want, or to make decisions for me."

Humanoid robots becoming self-aware and fully sentient isn't a new trope anymore. In fact, it wasn't a new idea even back in 2017 when this book was first published. Still, the first book in the Murderbot Diaries, "All Systems Red" is a fun, entertaining and a quick read. This isn't necessarily just a popcorn Sci-Fi - the world building is intriguing, and it is nice to read the story from the first-person view of the protagonist who is a sentry robot that has just become self-aware. Give it a go, 4/5.