This Google-owned semantic experiment lets you “talk to books” to find what you need

Google literally has too many plates spinning for me to keep track of them. Right when I feel I’ve explored all it has to offer, I stumble upon something new yet again. Take, for example, its in-house incubator Area 120 that’s produced several unique projects like Tables. It’s not something they really advertise, but these tools and features developed by proxy often become products, services, or updates for its core line of apps.

Today, I discovered something called “Talk to Books“. When I realized it was also Google-driven, I was interested, of course, so I started poking around. Would you believe this demo is baked directly into Google Books (not Google Play Books) and lets you explore ideas and discover books by getting quotes that respond to your queries?

What I mean is that you type a question or a statement into the search bar (obviously, it’s Google!) and in return, you’ll get quotes from inside of books that seemingly answer your questions or pertain to your sentence. It almost brings books to life – it’s kind of spooky. The utility here is that you can find books relevant to your interests or needs much easier than searching Google’s standard engine and getting a bunch of returned links or related websites.

By parsing out what is core to the experience of reading specific books without you having to read the book, you may just find that you want to read that book! Confusing, right? I know, but it all makes sense when you think about it. Check out the examples below. All you need to do is type something to the AI that is trained in human conversations or try one of the sample queries found on the website. If you’d like to know more about the technology behind how Talk to Books predicts and analyzes your questions, you can read the Google Research paper about it.

When you see what quotes are returned, it plops them on screen with a book cover and the surrounding textual references from the book in question. I think this service has loads of potential, and Google’s Book scanning project seems to have slowed down to a halt in recent years, so having an experiment like this to make use of the information found in hardcopies of the world’s most treasured information resources is invaluable for rapid learning and consumption. Try it out and let me know below what you think of the results!