1. My First Thoughts (Before the Research)

Honestly, as a sophomore in Electronics and Communication Engineering (ECE), I’ve mostly been using AI (like ChatGPT) to help explain concepts from my textbooks that feel like they’re written in another language. I figured that in the “real world,” AI is probably just doing the boring stuff, like running simulations for 10 hours or checking for wiring errors in a PCB layout. I didn’t think it was actually “creating” anything yet. I mostly saw it as a shortcut for students, not necessarily something that changes how a radio tower or a microchip actually works.

2. What the AI Claims (The Prompt)

I used Gemini to see what it thinks its own job is in my field.

  • Prompt: “I’m an ECE student taking Circuit Analysis. How is AI currently being used in electronics and communication engineering, and what’s coming in the next 5 years? Give me specific technical examples.”

The Output:

The highlights: It talked about “AI-driven EDA tools” (Electronic Design Automation) that can layout chips faster than humans, and “Self-Optimizing Networks” in telecom that fix signal drops before they happen. It sounded super futuristic, almost like engineers just sit back and watch the AI work.

3. Reality Check: Interviewing Dr. Yasser Gadallah

I sat down with Dr. Yasser Gadallah to see if the hype matched the reality. Since he’s my professor for Circuit Analysis I and an actual engineer, I knew he’d give it to me straight.

Here’s the breakdown of our talk:

  • I asked him if AI would eventually just “design” circuits for us. He kind of laughed and said that while AI is great at optimizing (making things smaller or faster), it doesn’t understand the “physical soul” of the circuit. If the AI places a component in a way that creates massive electromagnetic interference, it might not “realize” it until the board literally fails.
  • Telecom & Resource Management: He got more serious when talking about communications. He mentioned that AI is actually very real in managing how data moves through networks. It’s not just a buzzword there; it’s necessary because 5G and 6G are too complex for manual “if-then” rules.
  • Being a “Critical” Engineer: He warned me not to become a “button-pusher.” If I rely on AI to solve my circuits now, I won’t have the intuition to know when the AI is hallucinating a value that violates the laws of physics.

4. My Reflection:

After reading the “How to Argue with an AI Booster” article and talking to Dr. Yasser, my feelings are… mixed.

What surprised me: I was actually surprised by how much “hype” there is. The AI tool made it sound like I could just describe a circuit and it would be perfect. But Dr. Yasser made me realize that AI is basically just a very fast “guesser.” If you don’t know the math (KCL/KVL), you can’t check its work. It’s like using a calculator when you don’t know how to add, you’ll just believe whatever number it gives you, even if it’s impossible.

What scares/excites me:

  • The Scare: It’s a bit intimidating to think that a lot of the entry-level “grunt work” (like basic testing or routing) might be gone by the time I graduate. If the AI does the easy stuff, I have to be an expert at the hard stuff immediately.
  • The Excitement: On the flip side, I hate the tedious parts of circuit design. If AI can handle the boring optimization, I can focus on the actual architecture and innovation.

What I’m doing now: This assignment made me realize I can’t just “pass” my classes; I have to actually own the fundamentals. I’m also going to start learning more about how hardware and AI intersect (like Edge AI), because I don’t want to just use AI, I want to be the engineer who builds the chips that AI runs on. I need to be the one who can call out the “AI boosters” when they suggest something that physically won’t work.