“If I Were a Student Today, I’d Learn AI First” — What Jensen Huang Gets Right About Prompting

In a recent interview, Nvidia CEO Jensen Huang gave a piece of advice that should grab every engineer’s attention:

“The first thing I would do is to learn AI.”“Prompting AI is very similar [to asking good questions]. You can’t just randomly ask a bunch of questions. It takes expertise and artistry.”

If you’re using ChatGPT, Gemini, or any LLM to help draft reports, summarize specs, or build spreadsheet logic — you’ve probably seen it firsthand:

The quality of the output depends entirely on the prompt.

The Problem: AI Struggles Without Context

Many engineers treat AI like a search bar.

“Write me a safety report.” “Summarize this procedure.”

And then wonder why the answer is vague, overly wordy, or skips important technical points.

It’s not the model’s fault — it’s a lack of context.

AI doesn't know your format, your constraints, or what matters most to your work. That's why Huang compares prompting to a skill: it's not magic — it's communication.

The Solution: Prompt Like You’re Briefing a Smart Junior Engineer

Think of AI as a fast, intelligent assistant — but one that doesn’t know your process yet.

To get useful results, you need to:

• Break the task into steps

• Include the specific goal

• Provide any constraints or known data

• Reference an example format or output

  
  

When you do this, you’re not just asking AI for help — you’re guiding it to do useful work.

The Impact for Engineers

Engineers who learn how to prompt well see:

• Cleaner first drafts of reports or SOPs

• Better formatting for tables and summaries

• More useful spreadsheet formulas

• Less time spent rewording or rewriting AI output

  
  

Prompting isn’t about replacing your work — it’s about starting further ahead.

Want to learn how to write engineering prompts that actually work?Check out our hands-on course:🔗 AI for Engineers (Udemy)