Will you be an AI User as an Engineer or Scientist?

Will AI Revolutionize Engineering Jobs or Leave Engineers in the Dust?

If you had to decide whether your employees should learn to use AI, or ban them from using it, which would it be?

Like all new technologies, we tend to get emotional and polarised about whether it should be adopted or not. So why is this one, causing so much controversy?

Because it hits at the heart of white-collar jobs.

Previous technology has challenged blue-collar jobs – we won’t need truck drivers soon now that there are already driverless trucks operating. We won’t need shelf stackers, cashiers, bricklayers, the list goes on.

But a technology that sounds like it will replace engineers or scientists?

That sounds scary.

The other thing that sounds scary, is the speed the change is coming.

6 months ago, AI writing most of what you read on the internet was science fiction. Now it’s a matter of months away.

Is it months away from engineers not having to write any more reports because AI can do it for us?

Not exactly, but there will be a big change in what “work” looks like, very soon.

Will AI Change Engineering and Science Careers?

To answer this question, we need to consider what AI is.

It is not a super-fast human brain (yet), that can think divergently and convergently, creatively, and emotionally.

It is a tool.

And like every other tool that we have fashioned since first picking up rocks and sticks in the stone age, the tool is only as good as the hands that use it.

So just like every other time a new technology has surfaced, there will be an adjustment as the tool speeds up some parts of the job, allowing more time to create value in a different way.

Take CAD for example. When it became apparent that machines had the ability to create complex designs without human input (in the 1950s) when computer-aided design (CAD) programs were first developed, there were many draftsmen (because they were mostly men then) that were concerned they would be jobless.

But “automated design” did not lead to mass retrenchments.

CAD is still used, we still have draftspeople, we just have amazing 3D models and capabilities that people hadn’t dreamed of in the 1950s. Today it's possible for computers to generate plans based on user input such as geographical data or historical building plans. This allows designers to spend more time refining their creations rather than drawing them manually from scratch each time they need something new built.

AI in it many forms is just the next step of this automation.

There will be way less data analysis, report writing, menial data cleansing, manual quality checks, etc. There will be way more thinking and creativity to the roles.

So yes, AI will change what it looks like to be an engineer or a scientist.

But AI is similar to tools like hydraulic models, or 3D models, or any other “black box” process, and the same principles apply: input quality affects output quality. Garbage in = garbage out.

So no, you can’t expect a high school graduate to walk into the office and use AI to produce a quality engineering product or report.

Your job isn’t going to be replaced by AI.

BUT

And this one is a very big but.

You might just be left in the dust by someone who is skilled in using AI instead.

And that brings us back to the very first question - If you had to decide whether your employees should learn to use AI, or ban them from using it, which would it be?

If you want the competitive advantage, I recommend you get people skilled up as quickly as possible. So what skills do you need to use AI?

How to Prepare For AI-induced Changes to Engineering

To use AI effectively, you need two skills.

Skill #1: Feed the AI

Firstly, you must be able to feed the AI well. That is, you must understand your topic and the result you want, well enough to ask the right questions and input correct data.

This is the antidote to that black box challenge of garbage in = garbage out.

If you want a section of report written and you ask AI to write a report on “how changes in hydrology will impact the resilience of your wastewater infrastructure”, you are going to get a bland and shallow essay response.

If instead, you provide specific references to studies, reports or data sources, and you give research goals and objectives, you have a chance of getting something useable out of the tool.

Try asking it for “a report that helps me understand the impacts of changing hydrology on the performance or maintenance of wastewater treatment plants, and the potential for adaptation and resilience measures, in regional New South Wales; refer to the following sources…”

The output from this question will likely still need to be tweaked, but it can you well over halfway there.

Imagine how much quicker report writing could be!

Skill #2: Read the AI

Secondly, you must be able to read the AI well, and tell if you have a quality answer. You must have discernment to evaluate if the AI's output makes sense and is accurate.

We both work with young professionals, and the number one difference between the average professionals and the ones who will go far, is the ability to check their work and know when something is not right.

Too often people use a tool and expect because they used it that the answer is right. “Of course the pipeline can go there, the model said it could”. Never mind the mountain it would have to go through that makes it incredibly expensive to build.

Using AI is the same – you need the knowledge and experience as an engineer or scientist to test the output. To adjust the output to make sense, to adapt it to what is needed in the situation.

There's no point in using AI to write code for you, if you don't have the skills to troubleshoot it when it feeds you incomplete programs. The AI may have just switched the order of two lines, but if you don't have the ability to discern that, the whole program will be worthless to you. If you can see you just have to switch the lines back, then the AI has saved you loads of time because you aren't starting from scratch.

Remember, the AI answered the question you asked, and it may take multiple different questions to define the problem, so you will have to test the AI output multiple times and adapt your requests. You can’t do that if you don’t apply your expertise.

In professional settings such as engineering or science, don't fear using AI; it won't replace your job. However, someone proficient in AI might outpace you. Therefore, learning how to use AI early gives you an advantage.

H2 – Did AI write this blog?

If you are a discerning student, you should be asking at this point – so did you use AI to write this blog post?

Yes, of course I did.

But did I say, “write a blog on AI in Engineering”, press enter, post, and walk away?

No, of course not.

I used multiple AI tools, multiple different types of prompts, and mixed up all the results together.

I also actually wrote many of the words the old-fashioned way with the keyboard.

In fact, there are more words I wrote on the keyboard in the blog post than came from AI.

It seems I still have much to learn with using AI!