I Told My Students to Use ChatGPT for Their Coding Assignment. Here's What Happened.

Dialogue & Discovery

The Andraluma Compass - By Marco Lam

There's a common refrain among educators and employers today: "This new generation just uses AI to do their homework. They can't think for themselves." I hear this all the time. But in my IT class this week, with a group of bright 17 to 21-year-old students, I decided to challenge that assumption head-on.

The task I set them was, by traditional standards, impossible: create a functioning piece of software or a game by the end of their second term, despite having zero prior programming skills.

Then, I gave them the first instruction. Not to open a textbook, but to open an AI.

My rule was simple: Step 1 is to ask the AI to give you the code.

When I demonstrated this in class, using an AI to generate the code for a simple application in seconds, I could see the shock and wonder in their eyes. The room was buzzing. One student asked, "What was it like five years ago? How could students have done this?"

I gave them an honest answer. "Five years ago, students at your level would not have been able to finish an app like this in six months." Then I added the hard truth: "And that is why the job of a junior programmer—the person who simply translates a clear instruction into basic code—is going to disappear."

But this is where the real lesson began. I explained that I wasn't assessing their ability to copy and paste. The AI generating the code was just the starting point. The real assessment was based on three other steps:

1. Integration: Could they take the raw code and make it work in a real environment, like our class Microsoft Teams page?

2. Explanation: Could they stand up for three minutes and present their project, explaining what the code does? If I asked them a question about a specific function, could they find the answer by having another conversation with their AI?

3. Iteration: Could they take feedback from their peers and use it to have a more refined dialogue with their AI partner to improve the app?

The breakthrough moment came halfway through the class. A student wanted to change a feature in their AI-generated app. They were stuck. My answer was not to look at their screen or debug their code. My answer was: "Communicate with the AI."

For them, this was a revelation. The task instantly shifted from the technical problem of "coding" to the human skill of "dialogue." They had to think about their goal, articulate it clearly, and ask a series of smart questions to guide the AI to the solution. This was a completely new experience for them.

What I am teaching these students is not how to be a programmer in the 2010s sense. I am teaching them how to be an architect, a director, and a problem-solver in the 2020s. I am teaching them the skill of purposeful inquiry.

We are not raising a generation that can't think. We are on the cusp of a generation that will think differently. They won't be valued for the mechanical skill of writing flawless code from memory, but for their strategic skill of inquiry, integration, and communication with powerful AI tools. Our job as educators and leaders is not to ban these tools out of fear. It is to design new, more intelligent ways to assess and cultivate the real human skills that truly matter.

For Further Reading:

The ideas in this post are part of a global conversation about the future of education and work. These resources provide additional context.

1. The Global Perspective on AI in Education

• Source: UNESCO

• Report: https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research

• Connection: This guidance from UNESCO, the world's leading educational and scientific organization, provides a global framework for how educators should approach generative AI, reinforcing the need for new, human-centric teaching and assessment models.

2. The Future of Programming Careers

• Source: Harvard Business Review

• Article: https://hbr.org/2023/06/generative-ai-is-coming-for-the-programmers

• Connection: This article from HBR explores the profound impact that generative AI is having on the software development profession. It supports the post's core argument that the value of programmers is shifting from the mechanical act of writing code to the strategic acts of design, communication, and problem-solving.

3. The Shift to "Human Skills" in Tech

• Source: World Economic Forum

• Article: https://www.weforum.org/agenda/2023/05/future-of-jobs-2023-skills-outlook/

• Connection: This report on the "Future of Jobs" consistently highlights that the most in-demand skills are not purely technical. They are skills like analytical thinking, creative thinking, and technological literacy—exactly the "dialogue-based" skills that the assessment method in this article is designed to foster.