You've bought a great book.

You've finished reading it.

You highlighted many key points.

But three months later, when you really need it, you only remember:

“This book seems to have mentioned this.”

In the past, if you wanted AI to help you find the information, you had to:

Organize your notes yourself.

Scan the pages.

Find PDFs.

Copy and paste content.

Or simply ask AI:

“How does this book view this question?”

But the last method has a big problem:

The AI’s answer might not actually come from the book you bought.

On August 27, Google introduced Expert Intelligence to solve this problem.

Now, eligible Google Play Books eBooks can be directly added into Gemini Notebook.

Then, when you ask questions,

AI can answer based on:

The actual content of that book.

What exactly is Expert Intelligence?

It’s not another new chatbot.

It’s a cross-product initiative by Google.

The first step is integrating it into Gemini Notebook.

You could already add to Gemini Notebook:

PDFs.

Websites.

Google Docs.

Slides.

Images.

Audio.

And other data sources.

Now, there’s an additional source:

Specific eBooks you’ve purchased on Google Play Books.

As long as the book is eligible,

it can directly serve as a Notebook source.

No need to find separate files yourself.

No need to copy and paste the entire book.

Or to pre-organize the content.

What’s the difference from just asking Gemini "What does this book say?"

The difference is crucial.

Say you ask a general AI:

“How does Some Management Book suggest managers give negative feedback?”

The AI might base its answer on:

Public websites.

Book reviews.

Others’ summaries.

Its training data impressions.

Or even its own inference.

It compiles an answer for you.

That answer could be good,

but you might not know:

Is it really what the book says?

When you add the book you purchased to Gemini Notebook,

the answer is directly grounded in the book itself.

Simply put:

The answer cites that book as the source.

Gemini Notebook also provides source references for answers, letting you review the basis.

This is far more practical for learning,

research,

decision-making,

or citing the author’s viewpoints

than merely asking a generic chatbot.

It does more than just summarize a book

This is the easiest point to underestimate.

If it’s just:

“Give me five key points,”

any AI can do that already.

The real value of Expert Intelligence lies in:

Letting you keep working with that book.

For example, with a management book,

don’t just ask:

“What are the key points of this book?”

You can ask:

“According to the author’s method, what should I tackle first with my team’s current issue?”

Then add:

Your own meeting notes.

Work records.

Project plans.

All into the Notebook.

At this point, you’re combining:

The author’s framework,

plus

Your actual situation,

and letting AI compare the two.

Books can be used together with your own data

Google emphasizes this feature.

Gemini Notebook already allows multiple sources in one Notebook.

So now, instead of:

One book → One summary.

You can have:

A book.

Your notes.

A project plan.

Research data.

Some images,

all organized in the same Notebook.

For example,

You’re reading a brand strategy book,

and include your company’s branding materials.

Then ask:

“Based on this book’s framework, which three parts am I most clearly missing?”

This turns a reading tool

into a

work tool.

It can also turn your book into an Audio Overview

Gemini Notebook has a popular feature called Audio Overview.

Simply put:

AI can turn source data into content you can listen to.

Now, eligible eBooks can be a source for this as well.

So a book you already bought can be converted into:

Audio Overview,

Infographics,

Quizzes,

and other learning materials within the Notebook.

This is especially useful for content-heavy books that require repeated learning.

First time:

Normal reading.

Second time:

Ask questions.

Third time:

Listen to the Audio Overview.

Fourth time:

Take the quiz.

You’re not telling AI to “skip reading”.

Instead:

After reading, you use different ways to recall the knowledge.

Which books are eligible?

Not all books in Google Play Books qualify.

Google clearly states:

Only:

Specific eligible eBooks.

Currently, this mostly applies to eligible English eBooks.

When opening a book’s page in Google Play Books, if you see a “Tools” section listing Gemini Notebook, it means the book is eligible.

In Gemini Notebook,

under Add Sources, select Play Books

to see your books.

Eligible books can be added.

Books that don’t qualify may still appear on the list but will be unselectable.

So don’t assume:

“From today, all eBooks I purchased can be added to Gemini Notebook.”

Not yet.

Do EPUB files you uploaded yourself to Play Books count?

Currently, no.

Google Help clearly explains:

If you uploaded a book yourself to your Google Play Books Library,

it cannot be used as a Play Books source for this feature.

Expert Intelligence only works with:

Eligible books purchased through Google Play Books.

This is distinct from:

“I own an EPUB, so I uploaded it to Play Books.”

The rights are different.

Of course, if you legally own other file formats and Gemini Notebook supports that file type, you can still upload those as normal sources.

But that is a native Gemini Notebook feature,

not part of this Expert Intelligence Play Books integration.

Why does Google require that you must own the book?

This is one of the most important aspects of this feature.

If Google simply put all books on the market into AI,

the issues would be:

What about authors?

What about publishers?

Why can others access paid content for free?

So this time, Google has created a clear control for publishers and authors:

You must own the eligible book to use it inside the Notebook.

And if you share your Notebook with others,

but they don’t own the book,

sharing the Notebook won’t grant them the right to access the entire book.

Google restricts that source accordingly.

If they want to use it, they must purchase the book themselves.

This contrast interestingly echoes yesterday’s AI newsletter discussion on “How AI training data is sourced.”

One side has:

Unclear content origins leading to lawsuits.

The other side is trying to:

Embed content rights into AI features.

Can you add all your purchased books to one Notebook?

Technically, you can add multiple eligible sources.

But in practice, I don’t recommend creating a Notebook on day one called:

"All My Books"

and dropping in a hundred books.

The more sources,

the more you risk facing another problem:

Which book is actually answering?

A better method is to create Notebooks based on specific tasks.

For example:

“Pricing Strategy.”

Add two or three pricing-related books,

plus your own product data.

“Managerial Skills.”

Add management and communication books,

plus your meeting records.

“Content Strategy.”

Add content, branding, and marketing books,

plus your website data and planning materials.

This way, AI isn’t a library stuffed with books,

but rather a

workspace solving a specific problem.

Who is it suitable for?

First:

Students and self-learners.

After finishing a book, you can follow up on confusing points.

You can also turn content into:

Quizzes,

Reviews,

Audio Overviews,

Infographics.

Second:

Working professionals.

For a management, negotiation, or project management book, you don’t have to just stop at “I read it.”

You can add your current work issues into the Notebook and compare with the book’s approach.

Third:

Solo entrepreneurs and small businesses.

For example, with a pricing book,

add your services,

costs,

customer issues,

and current plans,

then let AI help find areas to rethink based on the source.

Fourth:

Content creators.

When you want to truly understand a book’s arguments,

you can ask the Notebook to:

Summarize the author’s original claims,

cite sources,

and then think independently,

instead of relying on secondhand summaries found online.

Will it make people read less?

If you just use it to ask:

“Summarize this 400-page book into 10 sentences,”

maybe.

Because you’re treating AI as a tool

to skip reading.

But if you use it to ask:

“Why does the author say this?”

“Under what conditions does this concept not apply?”

“Are the views in chapter 2 and chapter 7 contradictory?”

“Use this approach to check my current issue,”

then AI might help you use the book

more often.

The real value isn’t:

AI making you read 300 pages less,

but rather being able to recall an author’s knowledge

months after you read the book, as problems arise.

If you want to try today, just do one thing:

Don’t buy ten new books yet.

Open your Google Play Books library.

See:

Which book you’ve already purchased and truly want to make use of.

If it supports Expert Intelligence,

add it to Gemini Notebook.

And don’t ask:

“Summarize it.”

Instead, ask a real current problem:

“According to this book, what’s the first thing I should check to address my current issue?”

Then see which passage is cited in the answer.

At that point, AI is not replacing your reading.

It’s letting the knowledge you already paid for

come back into your work today.

Today, let’s improve together with AI.

Learn one AI skill daily.

Save a little time each day.

Enhance your abilities bit by bit.

SasaDaily, growing with you.

Recommended Reading

Today’s AI Tool|What is NotebookLM? Learn to Use AI to Organize PDFs, Notes & Meeting Records in 5 Minutes

AI Q&A|Will AI Peek at My Uploaded Documents? What You Must Know Before Using AI