A Bit about AI

I stayed at my sister’s place in Karawaci, near UPH (Universitas Pelita Harapan). As I passed by the university, I noticed they now have a campus dedicated to AI. AI is the big hype right now—but do people really understand how it can be applied meaningfully in daily life?

In many cases, I’ve met people in upper management who believe that everything can be done with AI, with minimal effort. But have they actually tried using AI to achieve their goals? And if so, what were the results? The real issue is that many don’t fully understand the technology, nor do they treat AI projects with the same rigorous testing and lifecycle management as they would with any new software.

Here’s an example: I thought about building predictive models using historical data (I took a course related to machine learning a few years ago, and I think this is not an easy subject to master). Here’s the question: instead of going deep into learning the methods, can we just let AI automatically generate predictions? If the answer were yes, then why would anyone bother learning at all? It would mean anyone could be anything with almost no effort. That’s exactly the misconception many people have about AI.

From my personal experience working with AI in the past few weeks, using it effectively requires a lot of trial and error. The answers often need validation, and sometimes AI outputs need to be trimmed or refined. Without enough knowledge or experience, you might not even recognize when the answer is wrong—and that’s when important details get missed.