I Asked AI to Pick My Next Read. Here’s How It Went

If you’re a book lover like me, you probably have a massive TBR list but still struggle to find your next read. Why not put the decision out of your hands entirely and ask an AI chatbot to choose your next book? I did, and the results were pretty interesting.

Can AI Chatbots Help Me Discover My New Favorite Book?

My current strategy for finding new books involves browsing recommended reading lists, scanning reviews on Goodreads, and checking out what’s trending on BookTok. However, the problem with these methods is that they are often hit or miss. Wouldn’t it be great if you could guarantee that you’d love every book you read from now on? That’s where AI comes in.

Many book catalog websites and online bookstores already use AI to suggest new reads based on your past purchases and reviews, but AI chatbots take things to a new level. Since you can converse with them, you can be more specific with your requests and describe exactly what book you want. Additionally, since these chatbots learn, they will offer better recommendations the more you use them—at least, that’s the idea. It remains to be seen whether this theory holds up to reality.

So, without further ado, let’s find out!

Setting the Baseline With Traditional Book Recommendation Pathways

We need something to compare the AI recommendations against, so, I went looking for books the old-fashioned way.

My first stop was Goodreads, where I went looking for a romance book. I focused on finding a recent release that would pull on my heartstrings but was still lighthearted enough to make me laugh and kick my feet. After several minutes spent looking through a dozen reviews, I finally settled on Funny Story by Emily Henry.

Next on the list was Barnes and Noble in search of a mystery book. Since I’m not very familiar with the genre, I put my trust in the editors and chose Three Inch Teeth by C.J. Box.

Lastly, I went looking on Amazon for a fantasy novel. I had to wade through the murk of ChatGPT-generated books, but after several minutes of browsing, I finally stumbled on The Honey Witch by Sydney J Shields.

Now that that’s settled, it’s time to explore the AI recommendations.

Exploring Romance Book Picks With ChatGPT

I asked ChatGPT to find me a romance book with a fake dating trope that was lighthearted and fun but that still had emotional depth. For guidance, I included similar books that I’d enjoyed in the past in my prompt and sent it off. ChatGPT responded with The Plus One by Mazey Eddings, a synopsis of the book, and a couple of reasons why it recommended this particular book.

I loved ChatGPT’s recommendation. It fit my criteria perfectly and the book seems interesting enough from the synopsis. However, I noticed that it was published in April 2023—not exactly a new release. In hindsight, this makes sense since I’m using the free version of ChatGPT, and it’s been trained on data only up to that date. Aside from that, it was a solid pick.

Gemini Has Some “Interesting” Murder Mystery Book Suggestions

When I asked Gemini for a newly released murder mystery book that would get my adrenaline pumping, it returned some Chinese text and an Amazon link to the book The 24th Hour: Is This The End? by James Patterson and Maxine Paetro. (What’s up with that Google?)

I wasn’t quite satisfied with this first result, and just in case my first query triggered some weird glitch, I gave Gemini a second chance.

I asked it to suggest a murder mystery book where the perpetrator intentionally leaves clues. Instead, it recommended Six of Crows by Leigh Bardugo, which was confusing since I’d read it, and it’s not a murder mystery at all.

In fact, it’s a young adult fantasy novel. Even Gemini acknowledged this in its response, where it said “This young adult fantasy novel is not your typical murder mystery.”

Poor showing, Gemini.

Discovering New Fantasy Reads With Copilot

I asked Copilot to recommend a book with magic and a strong romantic subplot. Copilot returned five titles, and the results were mixed. Two of the recommendations fit my requirements perfectly, while the other three did not align with my preferences at all. One suggestion was even a sci-fi book with no magic whatsoever.

Additionally, Copilot seemed to favor urban fantasy, although the reference books I provided were closer to high fantasy. Fortunately, when I asked it to choose one book, Copilot selected The Rage of Dragons by Evan Winter, which was one of the two titles that met my criteria.

Who Gave The Best Book Recommendation?

I loved ChatGPT’s romance recommendation more than the book I found on Goodreads. Copilot’s suggestion was decent, but I was more excited about the book from Barnes & Noble. As for Gemini, its recommendations were inconsistent, so Amazon took the win in that category.

I learned a couple of things from comparing the two ways of finding new books. The first was that asking a chatbot is the fastest way to find what to read. I never realized how much time I spent deliberating over books on Goodreads until now. The platform has so many diverging opinions, that it can be difficult to form yours about a book.

AI chatbots, on the other hand, provide a definite answer without the decision fatigue that comes from sifting through countless reviews.

Another thing I learned is that AI chatbots are great at singling out specific themes in books. With Amazon and Barnes & Noble, the recommendations were all over the place, and although you can find some curated lists, they don’t always exist for really niche requests. AI chatbots, on the other hand, can quickly comb through reviews and find exactly what you’re looking for.

Finally, I was reminded that AI chatbots still get things wrong from time to time. Gemini recommending a YA fantasy book when I asked for a murder mystery is a perfect example of this.

So, even though the traditional recommendations technically won this battle, I will be incorporating AI into my book selection process. I might use it to choose between books when I’m undecided or when I’m trying to find something particularly niche. But I’ll still visit Goodreads occasionally to read a couple of reviews and get a feel for what other readers think before diving into a new book.