This blog is taken from The Real Work magazine | Edition 4 | coming in December 2026. The Real Work is a print magazine about work, culture, and the decisions that shape how people actually experience their jobs. Discover it here.
A few months ago, an article on the front page of our local paper jumped out at me. A mining operator was replacing full crews of dump truck operators with autonomous trucks.
It wasn’t the technology itself that caught my attention. Autonomous mining equipment isn’t a new concept. Driverless trucks have been operating across Australian mine sites for years, and BHP has been running autonomous trains in the Pilbara since 2019.
What made this article land differently was something I’d read only a few hours earlier.
A thoughtful LinkedIn post — and an even more thoughtful comment thread — was discussing the entry-level HR tasks now being automated by AI. The administrative work, basic research, first drafts, data entry and routine coordination graduates once did while finding their feet in the profession.
The concern wasn’t simply that these tasks would disappear.
It was what happens when we remove the work through which people develop the knowledge expected of them later.
Entry-level work isn’t just low-level work
There’s a tendency to look at entry-level tasks as the work people must get through before they’re trusted with anything important.
But those roles do something far more valuable.
They ground us.
They teach us how the workplace operates, how decisions connect and why a process exists in the first place. They expose us to mistakes, exceptions, competing priorities and the distinctly human reality that the answer written in a policy isn’t always the answer that will work in practice.
In HR, you learn by preparing the contract, sitting in on the meeting, listening to how an experienced practitioner asks a difficult question and watching what happens when a technically correct process lands badly with an employee.
You learn that knowing the legislation and knowing how to apply it are two very different capabilities.
That depth isn’t downloaded with a degree. It’s built through exposure, practice, feedback and time.
If we automate the work that creates those foundations, but continue expecting graduates to enter roles requiring the judgement those foundations once produced, we haven’t solved a capability problem.
We’ve created one.
The mining career funnel
This is where the autonomous truck article brought the issue into sharper focus.
Mining machinery operators have traditionally had a reasonably clear career pathway. Whether working open pit or underground, you can picture it as a funnel.
At the wide opening are the truck operators — the truckies. For many people, this is the starting point of an operating career.
From there, operators build experience across the site. They learn how the operation fits together, develop an understanding of the equipment and demonstrate whether they have the judgement, capability and temperament to progress.
As the funnel narrows, theoretically at least 😉, the strongest operators move into increasingly skilled positions.
At the pointy end are roles such as digger operators in open pits and jumbo operators underground. These positions require considerably more than knowing how to operate a machine. Underground in particular, experience tells an operator what the creaks, movement and sounds coming from the walls may be telling them.
That judgement has been built over years.
These people didn’t arrive at the narrow end of the funnel by skipping everything beneath it. They started at the opening and accumulated knowledge of the equipment, environment, risks and operation as they progressed.
Now picture what happens when we cut a hole halfway down the funnel and begin feeding people through there instead.
Where did they develop the knowledge that once came from the stages below?
Who taught them how the whole operation connects?
How do they build judgement without the experience that previously shaped it?
And who carries the risk while they’re learning?
This isn’t an argument against AI
AI and automation can remove repetitive work, improve safety, increase consistency and give people more time to focus on higher-value activities. Autonomous trucks may also remove workers from hazardous environments and address operational challenges that have existed in mining for decades.
The technology isn’t necessarily the problem.
The problem is assuming we can remove the beginning of a career pathway without redesigning how people develop the capability needed for its later stages.
We can’t automate foundational work and then act surprised when emerging employees lack foundational knowledge.
We also can’t expect experienced employees to absorb the responsibility for accelerating that development without giving them the time, structure and coaching capability to do it well.
“Ask Jenny if you get stuck” isn’t a development strategy.
Neither is giving a graduate access to an AI tool and expecting them to assess whether its answer is correct when they haven’t yet developed the knowledge required to make that judgement.
AI can provide an answer. It can explain legislation, draft a document or identify a pattern in data.
It can’t give someone the workplace context that comes from seeing how a decision affects payroll, operations, employee trust, leadership credibility and business risk all at once.
That understanding has to be developed.
What does your organisational funnel look like?
Think about the senior, technical and specialist roles within your own organisation.
Where have the people currently performing those roles traditionally started?
What work gave them their base knowledge?
Which experiences taught them how the organisation operates beyond the boundaries of their own position?
Now consider the roles, tasks and opportunities being removed through automation, outsourcing or restructuring.
Are they merely tasks?
Or are they part of the pathway through which your future capability has always been built?
Because if those entry points disappear, organisations will need to replace them with something deliberate: structured development pathways, meaningful supervision, job rotations, coaching, practical exposure and room for emerging employees to make mistakes safely.
That requires investment.
It also requires experienced people who know how to explain not only what needs to be done, but why. People who can recognise the gaps created when someone arrives halfway through the funnel and help them build the knowledge they haven’t had the opportunity to acquire.
Without that, we risk producing people who can complete higher-level tasks but don’t understand the foundations beneath them.
And eventually, we run out of experienced people available to teach them.
The conversation about AI replacing entry-level work can’t be limited to efficiency, cost savings or the number of roles removed. It must also consider workforce capability, career pathways and where the next generation of experienced employees will come from.
Removing the starting point doesn’t remove the need for everything people once learnt there.
It simply leaves us to discover, usually much later, exactly what went missing.