Will Every Child Have a Personal AI Teacher?

(5 minute read)

Imagine walking into a classroom in 1776. Books were expensive and scarce, many children had none of their own, and knowledge travelled at roughly the speed of a horse or sailing ship. If the information you wanted was sitting in a library two hundred miles away, it might as well have been on another planet.

Now imagine a classroom in 2076. A twelve-year-old girl in Britain is studying astronomy alongside children in India, Brazil, Japan and Nigeria. They are all speaking their own languages, yet each hears everyone else perfectly in whichever language they prefer.

Working with every student is their own permanently assigned artificial intelligence tutor that knows exactly what they understand, where they are struggling and how best to explain the next idea. One child learns visually, another by solving problems and another through discussion, so each receives a slightly different lesson even though they are studying the same subject.

Education has already changed enormously during my lifetime. When I was young, finding something out meant asking somebody who knew, looking it up in an encyclopedia or going to a library.

Today, almost everything humanity has discovered is available within seconds on a device we carry in our pockets. Yet the basic structure of education has changed far less. Thirty children still often sit in a room while one teacher explains roughly the same subject to all of them at roughly the same speed. The cleverest may become bored, those struggling can fall behind, and the teacher has to pitch the lesson somewhere in the middle.

Artificial intelligence would change that. Wealthy families have always been able to hire private tutors who adapt lessons to an individual child, but no national education system could provide that level of attention to everyone. By 2076, perhaps every child could effectively have their own private tutor working alongside the human teacher.

Imagine a child struggling with algebra. Her AI tutor notices that she understands equations until fractions appear, so it tries another explanation, then a visual example, a game or a practical demonstration. Another student might race through the same lesson and immediately move on to something harder. Neither child has to wait for the other, which could transform education from something standardized into something genuinely personalized.

Geography might matter far less as well. A brilliant mathematics teacher in Italy could teach students around the world simultaneously, while a child in a remote African village attends the same physics course as someone at an expensive private school in London. Instant translation could allow students from dozens of countries to study together while each hears the lesson in their own language.

Virtual classrooms could make lessons far more immersive. Students studying ancient Rome might walk through a reconstruction of the Forum, an astronomy class could stand on the surface of Mars, and biology students could travel through the human bloodstream, watching white blood cells attacking life-threatening viruses.

That raises another question. If information is permanently available, how much do we actually need to remember? Schools have traditionally placed enormous emphasis on memorization because knowledge used to be difficult to obtain.

That does not mean knowledge becomes unnecessary. You cannot think intelligently about history if you know nothing about history, nor judge whether an answer is sensible if you have no understanding of the subject. But education may place less emphasis on remembering facts and much more on knowing what to do with them.

Critical thinking, creativity and judgment could therefore become more important. If a machine can produce a competent essay on the causes of the First World War in seconds, asking thirty students to write essentially the same essay becomes a strange way of testing them. A better exercise might be to give them several conflicting explanations and ask which one they think is strongest and why.

There is another possibility that takes us further into science fiction. We already use caffeine to make ourselves more alert and medicines to treat conditions that affect concentration. Over the next fifty years, perhaps treatments will improve memory, concentration or learning. If a safe medicine allowed a student to remember twice as much, would taking it be cheating? Brain-computer interfaces could take that further, perhaps allowing people to search for information mentally.

There is, however, a problem with all of this that has little to do with technology. Children do not go to school simply to absorb information. They need other children. They need to play, argue, make friends, fall out, make up again, learn how to share, compete, cooperate and understand the hundreds of social cues that adults eventually use almost without thinking.

Even home-schooled children often meet through play dates, sports, clubs and group lessons because parents understand that children need regular contact with other children.

That is why I doubt the school of 2076 will consist of millions of children sitting alone at home with AI tutors. It may instead become a more social place, not a less social one.

AI could handle much of the individualized academic teaching while human teachers spend more time helping children work together, debate ideas, solve problems in groups, play sport and learn how to behave as members of a community.

Human teachers may matter for another reason too. A computer might recognize that a child repeatedly gets a mathematics question wrong. A good teacher may realize that the child's parents are divorcing, that she barely slept the previous night and that mathematics is not really the problem. Teachers encourage children, recognize potential and sometimes persuade them that they can accomplish something they assumed was beyond them.

There is also a darker danger in relying too heavily on machines. If children ask an AI whenever they encounter a difficult problem, will they develop the persistence required to solve things independently?

More importantly, what happens when the machine is wrong? A confidently delivered falsehood can be much harder to recognize than obvious nonsense. A child in 1776 faced a shortage of information. A child in 2076 may face the opposite problem: far too much of it.

Education may therefore become less about finding information and more about deciding what deserves to be trusted.

There are also questions about inequality. If every child receives access to an excellent AI tutor, technology could reduce educational inequality on a remarkable scale. A poor child could potentially receive teaching every bit as sophisticated as a billionaire's child.

But wealthier families may buy better systems and some governments may restrict information, so technology will not automatically make education fairer.

I suspect the most successful classroom of 2076 will combine the strengths of both humans and machines. AI could provide individual tuition, instant translation and access to almost everything humanity has learned, while human teachers provide judgment, encouragement, discipline, empathy and leadership.

Children would still spend time together because education is partly about becoming a functioning human being rather than merely an efficient processor of information.

The skills schools value may change as well. For much of history, education rewarded people for remembering information and following instructions accurately. The economy of 2076 may value creativity, curiosity, adaptability, judgment, the ability to recognize nonsense and the ability to work with people who think differently.

Education may also become something we continue throughout life. If people live longer and change careers repeatedly, a seventy-year-old might return to university to learn an entirely new profession, supported by an AI tutor that already understands what she knows and adjusts the course accordingly.

There is something wonderfully optimistic about that possibility. In 1776, access to knowledge was largely determined by where you were born, how wealthy your parents were and whether somebody possessed the right books. By 2076, a child almost anywhere on Earth may have access to more knowledge than the greatest scholars in history could have imagined.

As a science fiction writer, I find that fascinating. Humanity has always advanced because we learned, then taught one another, and each generation inherited the discoveries of the one before it. Perhaps AI will simply accelerate something we have been doing for thousands of years.

For the first time, every child may have almost all of human knowledge sitting beside them, patiently waiting for their next question. The challenge will be making sure they still learn how to think for themselves, and how to live with everybody else.

What do you think? Would you have learned more if you had grown up with a personal AI teacher, or is there something about the traditional classroom that we should be careful not to lose?

Till next time,

Brian

BA Gillies

*******

If you've been meaning to dive into the Return to the Galaxy Universe, now’s your chance!

Whether you're a new reader or just haven’t grabbed your copies yet, now’s the time to catch up on the award-winning series readers are calling “better than Scalzi” and “the best book since Heinlein died.”

If you haven’t started the series yet, Book 1 is still available to buy for 99 cents or read for free on Kindle Unlimited:

Read Return to the Galaxy on Amazon UK

Read Return to the Galaxy on Amazon US

*******

Discover More Sci-Fi Adventures

Each week, I’ll team up with a group of talented sci-fi authors to bring you an incredible selection of books. Whether you’re looking for thrilling space battles, deep space mysteries, or first-contact encounters, these books are packed with stories to fuel your imagination.

Author Spotlight: