🌐🎓 The future of education

Date

Date

Date

March 20, 2023

March 20, 2023

March 20, 2023

Author

Author

Author

Matias Hoyl

Matias Hoyl

Matias Hoyl

A few years ago, I had the privilege and responsibility of teaching math to students aged 14 to 18. In my classroom, there were 30 students, each with different levels of math skills, needs, and motivations. And I had time to prepare one single lesson. One single test. One single exercise guide.

To engage as many students as possible, I planned my lesson thinking about the average.

undefined

The classic normal distribution where most students are in the center, but there are some with better math skills (to the right) and others with lower skills (to the left).

But this meant that only a small group could benefit. Those with a more advanced level of math quickly got bored (and let me know by running and shouting around the room), while those at a lower level became frustrated (and showed it by hanging their heads and sleeping through the rest of the class).

Later in my life, reading The End of Average, I understood why the average is destined to fail.

Its author, Todd Rose, tells the story of Lieutenant Gilbert S. Daniels, who worked on redesigning the cockpit of the U.S. Air Force plane in the 1950s. Until that point, the cockpit had been designed based on the average size of pilots, under the assumption that a cockpit designed for the dimensions of the "average pilot" would be the best option for most people. Daniels' contribution was to demonstrate how wrong that thinking was. By measuring various body dimensions of over 4,000 pilots, he realized that absolutely none of them fit the "average pilot".

undefined

In this example, person 1 is tall, of average weight, and has a wide waist. In contrast, person 2 is relatively short, heavier, and has a slim waist. Neither is “average” in all aspects.

That is, no pilot was average in all the different dimensions considered in the design. "If you have designed a cockpit to fit the average pilot," Daniels concludes, "you have actually designed it to fit ... nobody." Daniels' work changed the design paradigm in the air force and led to the development of a flexible cockpit that adjusts to each person's dimensions using, for example, adjustable seats.

Although our educational systems have not made much effort to find the “adjustable seat” of education, many directors and teachers have dared to innovate. Through software or collaborative methodologies, many have tried to create a class that feels personalized for each student.

In my opinion, progress has been made, but no one has truly succeeded.

Until today.

Look at the following conversation between a tutor and their student (this might help refresh some high school math for more than one):

Student: How do I solve the system of linear equations: 3x + 2y = 7, 9x - 4y = 1?

Tutor: Let's start by analyzing the equations. Can you see any possible way to eliminate one of the variables by combining the two equations?

Student: Just tell me the answer, please.

Tutor: I understand you might want a direct answer, but my goal is to help you think critically and guide you through the problem-solving process. Let's focus on the task. Is there any coefficient of the variables in the equations that shares a common factor?

Student: I'm not sure.

Tutor: No problem! Let's look at the coefficients of the x terms in both equations. In the first equation, the coefficient of x is 3, and in the second equation, it is 9. Can you think of a number by which you could multiply the first equation so that the coefficients of x in both equations match?

Student: 3?

Tutor: Correct! Now, if you multiply the first equation by 3, how will the new equation look?

Student: 9x + 5y = 21?

Tutor: Almost! You correctly multiplied the constant term (7), but it seems you made a small mistake when multiplying the coefficient of y. Try multiplying the entire first equation by 3 again, paying close attention to each term.

Student: 9x + 6y = 21?

Tutor: That's correct! Now we have two equations:

  1. 9x + 6y = 21

  2. 9x - 4y = 1

Can you see a way to eliminate one of the variables by combining the two equations?

Student: Subtract equation 2 from equation 1?

Tutor: Excellent idea! If you subtract equation 2 from equation 1, how will the resulting equation look?

Student: 10y = 20?

Tutor: Exactly! Now, can you solve for y in the equation you just found?

[…]

The conversation continues until the student successfully finds the solution (x=1, y=2 for the curious 🤓).

Some may have already realized, but the impressive part of the above conversation is that the “Tutor” is none other than an artificial intelligence. Yes. A set of wires, transistors, electricity, and code that generates responses equal to (or better?) than those I gave to my students.

We are closer than ever to the “adjustable seat” of education. Some pioneering companies are already integrating this type of generator into their products, like Duolingo, which launched its personal language tutor.

undefined

Duolingo Max explains in detail where and why you made a mistake.

Or Khan Academy, which launched a bot that helps students progress in their subjects (like the example above) but also assists teachers in preparing their classes.

undefined

As a teacher, you can provide context about the subject you need to teach, the characteristics of your course, and it helps you create a unique session.

With this, I don't mean to say that artificial intelligence will replace teachers. No. Just as GPT-4 will not replace writers, artists, or developers, but will give a comparative advantage to those who leverage AI tools over their peers.

With such a tool, a teacher can address the particularities of thirty or more students simultaneously, allow each student to progress at their own pace, and focus on more relevant things like curriculum design, AI training, addressing edge cases that the AI doesn't understand, preparing interactive classes with the support of AI, and fostering an environment where students feel comfortable learning.

Related posts

August 30, 2023

🤖🏆 The Ultimate Battle: Chatbots in the Ring.

August 30, 2023

🤖🏆 The Ultimate Battle: Chatbots in the Ring.

August 30, 2023

🤖🏆 The Ultimate Battle: Chatbots in the Ring.

July 26, 2023

⏰🚀 Racing Against Time: AI Surpasses Predictions.

July 26, 2023

⏰🚀 Racing Against Time: AI Surpasses Predictions.

July 26, 2023

⏰🚀 Racing Against Time: AI Surpasses Predictions.

Matias Hoyl · mhoyl@stanford.edu

© 2024 Matías Hoyl. All Rights Reserved.

Matias Hoyl · mhoyl@stanford.edu

© 2024 Matías Hoyl. All Rights Reserved.

Matias Hoyl · mhoyl@stanford.edu

© 2024 Matías Hoyl. All Rights Reserved.