The End of Books? The Future of Reading in the Age of AI Magic

Napoleon is supposed to have said, Show me a family of readers, and I will show you the people who move the world. People tend to unconsciously recognize how powerful the activity called reading is. For almost 600 years now, the printed book, and its spiritual successor the eBook have been the chief helper in kindling and sustaining the fire of enlightenment throughout the world. Many great scholars and erudites have shared their love and pride towards their personal libraries. Such history has built up the aura of the written word.

Unsurprisingly, given the established power of written works, a large number of people have always been concerned about benefiting properly from reading. On any given day on the web, we see a good amount of online activity around people trying to get better at things. People go to various online avenues, and ask - how do I get better at reading? How do I make it a habit? How do I read more? How do I remember better? What should I read? And so on. Maybe you too have asked such questions once, or maybe you are asking such questions today (like me).

Either way, it is circa 2023, and we are living through a special period in human history, the time when we are privileged to see the first flickers of AGI (Artificial General Intelligence) in action. It is an amazing time to be alive, and we are lucky to witness the emergence of such magic in the world. At the same time, we are forced to ask some hard questions about what’s in store as to the future of reading, in the age of AI magic. Is it the beginning of the end, of the great tradition of written works? Are all the questions we are asking about reading now irrelevant? Are we asking the wrong questions?

I will present my analysis in the following way:

• First, I’ll paint a zoomed-out picture of how humanity reached where it is today. In particular, I highlight the rate of change in terms of complexity in new significant happenings. The rate of change has an implication to what the future holds.
• Second, I’ll try to present what the future of knowledge distribution and acquisition may look like. I’ll use the analogy of the hybrid model of personal and cloud computing working together in recent years to project the possibilities with AI.
• Finally, I will outline what sort of adaptation is required for the upcoming decade, and what are the costs of not making the necessary adaptation.

The path taken till now

The diagram above highlights some significant events in human history. Humans appeared around 2,00,000 years ago on the planet. It is perhaps a bit of a surprise that we didn’t think of the idea of writing till 5,000 years ago. Therefore, for almost 97.5% of human existence, we haven’t had the slightest idea of writing.

Humans became serious in terms of writing only 3000 years ago when we figured that the sounds of spoken words could be represented through symbols and formalized for efficient transfer of meaning. However, notice in the graph above that interesting things start happening immediately after the emergence of the writing idea. The printing press was invented merely 600 years ago by Gutenberg. It was a revolution in the distribution of knowledge, enabling a large swath of humans to acquire advanced perspectives efficiently.

About 50 years back, the silicon revolution happened, along with the Internet, and eBooks, increasing the overall efficiency of learning and communication of ideas across the world. Within a really short period, we have intelligent machines, the so-called artificial intelligence performing important roles in the day-to-day operations of running civilization.

The undeniable conclusion is that the rate of evolution for humanity has exploded, and if the past rate of growth is of any relevance, we should expect newer breakthroughs to happen within shorter durations of time. The slogan "Change is the only constant" is more appropriate today than ever, and it will be even more appropriate in the days to come.

As to the question of the future of reading books, given the rate of change, and improvements, we definitely must expect rapid changes in the way knowledge is created, distributed, and used. Since we know that books have had a solid run of about 600 years, we may give them at most about 100 years to completely become antique pieces to be preserved, but we must definitely expect their importance to go down over time, even in the medium term. It may well happen that the book as a mainstream knowledge acquisition method will end up practically dead within 25-30 years.

What’s in store for the future of reading?

The most rigid part opposing change is the habits of the older people, decades of knowledge acquisition methods, and deeply entrenched teaching methods in the schools and colleges. We are awaiting an influx of pain and confusion as the old world must adapt to make way for the new world. However, I suspect the younger people will have no qualms about skipping reading cover to cover, and simply getting their information through AI.

Moreover, whether we like it or not, books are going away in the long term. The way I see it, the core value proposition of books is that they aid with the survival and thriving of civilization. In one potential configuration of the future, AIs could act as a translators of ideas from one mind to another.

In the long view, one can imagine personal AI helping with communicating with the world and learning things from the world. AI is going to help us be more effective in the world. There might be more powerful general AIs around aiding personal AIs as well.

I imagine a hybrid of small and personal AIs, learning about the preferences of the individual users, connected to powerful AIs operating on a larger or global scale. The reason to bet on a hybrid AI ecosystem is the evolution of computing itself. At different times of computing evolution, either personal computing or large-scale cloud computing (not both) has been of prime importance. However, over time we see a combination of personal and cloud infrastructure working together.

I think a model like this can enable some powerful improvements on the medium of books:

1. Presently a strong advantage of the book is the immersion it allows in the author's world. With personal AIs, it may become possible to convey a deeper immersion in the author's world. Your personal AI could grab all the nuances from the author's AI, and help you "acclimatize" to it in a progressive and guided way. Especially for the works of philosophers, it might be possible to send out a debating partner in the voice of the greats to engage the reader whenever needed.
2. AI is a more dynamic "medium" compared to books. So, AI can still behave like a book, by being static and immutable (while the book cannot pretend to be an AI!). So AI is a super powerful capability layer on top of static media such as books. It is a superset, an extension of older media.
3. Books have helped preserve, enhance and distribute culture across the globe. However, ultimately we must recognize that culture is a living reality within human minds, and such state is achieved only through training. AI can perform a more active role in training people, compared to static books.

The 10-year view

The meaning of “fluent reader” has changed now. Cover-to-cover reading is dead. There is value in synthesizing knowledge creatively, to put things together in new interesting, and useful ways. And we can do it much faster than before. We can also do it better than before.

However, the minimum bar for knowledge acquisition and attainment has increased significantly. The level of knowledge that made someone an educated person will no longer make the cut. Given powerful AI tooling, a person is supposed to be 10x more capable, at synthesizing, validating, and questioning. They are expected to be able to deal with orders of magnitudes more the amount of information than before.

One hope with great AI tooling is that more generalized thinking will become possible for motivated human beings. The curse of specialization need not limit the human being as much as it used to. It becomes feasible through AI support to master way more fields now.

So, in the next 5-10 years, at the least, the following adjustments must be made for the individual to thrive in the new environment:

1. Cover-to-cover reading becomes more irrelevant with every passing day. AI can summarise sections, it can reformulate information to fit your needs, and can even synthesize new information from multiple works.
2. In the AI-enabled world, true creative and coherent work becomes more valuable, and mediocre work will probably get any recognition at all. Facts are a dime a dozen now, but what is still needed is arranging the facts in the most coherent and useful ways.
3. Verification of information is 10x more valuable. Skepticism is going to become extremely important in many areas. The market is going to pay more for better vetting of information, and also dealing with new types of systemic errors produced due to AI-enabled work.
4. The emphasis placed on memory during education in the past will work against you. Don’t put too much effort into trying to remember. You just need to remember “pointers” to concepts, which in turn helps you nurture a large web of pointers in a way that is not burdensome. So, a larger field of awareness is the future, more than a crisp memory of smaller amounts of information.

Costs of not adapting

Books are on their way to death; it may take a while, but it is happening. Those who do not adapt to the new realities will end up lagging behind. Your peers who adapt will become capable of doing way more with way less, while you will struggle to get a little done with a lot of effort.

Especially for the older folk, a rigorous effort is needed to unlearn older habits, such as excessive preference for memorization, cover-to-cover reading, accepting information without questioning, and emphasis on regurgitation of facts (rather than assembling them in creative ways).

The younger folk have an immense opportunity to rebel against the status quo and achieve knowledge levels never seen before for a human being. We are yet to witness the New Scholar, a person who has attained enormous knowledge through significant AI support. You could potentially be a real generalist, knowing many fields, thinking creatively, and reaching conclusions of the kind that was impossible before. If you don’t capitalize on it, you may come to regret it.