Students are told two contradictory things. One is that AI will remove entry-level technical jobs. The other is that nothing important has changed. Neither is accurate, and the gap between them is where you actually have to plan a career.

Where AI sits in an engineering workflowJudgement — what problem is worth solvingDomain knowledge — is this answer physically sane?AI assistance — drafts, code, summaries, first passesTooling — CAD, compilers, simulators, instrumentsAI moves the floor up. It does not move the ceiling down.
The honest position sits between the two loudest claims.

What has changed

The cost of producing a first draft has collapsed. A first pass at code, a summary of a long document, a boilerplate report, a routine calculation: these are now fast for anyone with access to a capable model.

The consequence is that tasks which used to be a junior person's whole job are now a small part of it. Being able to produce competent routine output is no longer, by itself, a job.

What has not changed

Knowing whether the output is right. This is the whole of it, and it is not a small thing. A model will produce a confident answer that is subtly wrong, and the ability to notice is built on exactly the fundamentals that feel slow and unglamorous to learn.

Also unchanged: responsibility. Somebody signs off. Somebody is accountable when a structure, a circuit or a system fails. That role requires a person who understands the domain.

What this means for how you study

The temptation is to use AI to get through coursework faster. This is the one use that damages you, because coursework is where the judgement gets built. Using a model to finish an assignment you did not understand converts a learning opportunity into a submitted file.

A better pattern is to attempt the work first, then use the model as a reviewer: ask it where your reasoning is weak, what you have not considered, what a stronger answer would include. That direction of use builds judgement rather than substituting for it.

The skill that rose in value

Specification. Knowing what to ask for, in enough detail that the answer is useful, and being able to tell a good answer from a plausible one. That is a domain skill, not a prompt trick, and it is learned by understanding the subject.

What to take away

  • First drafts got cheap. Judgement about whether a draft is correct did not.
  • Using AI to finish work you do not understand trades your training for a submission.
  • Attempt first, then use the model as a reviewer of your reasoning.
  • Specification and verification are now the high-value skills, and both are domain knowledge.

Thinking about engineering at Maria?

Our admission office can talk you through branches, eligibility and the counselling process for 2026–27. No obligation.