Geilton Xavier Blog

Senior Software Engineer • Cloud Native Developer • Tech Leader

A few days ago, I was listening to the podcast IA sobre Controle (I do not remember the exact episode). They talked about the use of AI in universities, its impact on education, if it helps or not, etc.

They shared this article: The Impact of AI on Computer Science Education, and I will summarize it for you.

Basically, a professor at MIT divided a class into groups of three to study the Fortran language:

Group 1 could use Google.

Group 2 could use a Llama model specialized in code.

Group 3 could use ChatGPT and similar tools.

The results were like this: The group that used ChatGPT and similar tools got first place. The group that used Llama had a similar score. The group with the lowest score was the one that used Google.

Then, they did another study, this time about retention. They gave the same three groups a new test, but this time without any external help, from AI or other sources. The result? The groups that used AI got the worst scores, while the group that used Google got the best score.

In short, it is clear that those who researched by themselves retained knowledge better.

My opinion is: AI is here to help. It can be a “friend” or a copilot. But do not give your intelligence to AI. Use it and learn from it. Sometimes (or always, haha), correct it. This way, you increase your productivity, keep learning, and stay critical – which is essential to remain relevant in any field.