In my work with organisations adopting AI, I’ve seen a growing focus on efficiency without equal consideration for long-term talent development. I wrote this piece to highlight the overlooked risk of eroding entry-level learning pathways—and to encourage leaders to intentionally design how human judgment and capability are built in an AI-enabled workforce.
The number everyone is quoting — and the one they should be
The World Economic Forum’s Future of Jobs Report projects that 170 million new roles will be created globally between now and 2030, while 92 million will be displaced. Most commentary stops there. Net positive — problem solved.
But that framing misses the most consequential part of what’s happening.
The 92 million roles being displaced are not evenly distributed across the workforce. They are disproportionately concentrated at the entry level — data entry, basic analysis, routine compliance checking, standard customer enquiries, administrative coordination. Precisely the work that has, for generations, served as the training ground for every professional who went on to lead.
Automate away entry-level work, and you don’t just change what people do. You destroy the pathway through which people learn to think.
The pipeline problem nobody is discussing
I’ve spent more than a decade delivering AI and automation implementations across financial services, utilities, logistics, and government in Australia and New Zealand. And the conversation I keep not hearing — in boardrooms, at conferences, in the trade press — is the one about what happens to the talent pipeline.
Not five years from now. Ten years from now, when today’s graduate intake should be stepping into senior roles, drawing on the practical foundation they built in their early careers.
If that foundation never formed, what exactly are they stepping into leadership with?
This isn’t a hypothetical. It’s a structural risk that is accumulating quietly in organisations that are correctly pursuing efficiency gains from AI, while neglecting to ask what those gains will cost them a decade down the line.
Where the risk is most concentrated
Consider financial services — a sector where I work extensively across ANZ.
Graduate analysts have traditionally spent their first two to three years doing work that, from the outside, looks routine. Building financial models. Processing documentation. Conducting due diligence. Running standard compliance checks. It is repetitive. It is detailed. And it is irreplaceable as a developmental experience.
That work teaches commercial judgment — the ability to sense when a number feels wrong, when a document raises more questions than it answers, when the technically correct answer is not the right answer for this client in this context. It builds attention to detail as a habit. It develops sector knowledge that cannot be acquired in a classroom. And it creates the pattern recognition that experienced professionals rely on in every high-stakes decision they make.
AI can now perform many of these tasks faster and more consistently than a junior analyst. That is not an argument against deploying AI. It is an argument for being deliberate about what is lost when you do, and how you rebuild it.
The same dynamic is playing out across professional services, consulting, law, accounting, and technology. The routine cognitive work that AI handles most effectively is precisely the work through which professionals have always developed their foundational expertise.
What a missing generation looks like
Imagine a mid-sized financial services firm in Auckland that aggressively automates its graduate analyst function in 2025 and 2026. The efficiency gains are real. The cost savings show up on the P&L. Leadership is pleased.
Fast forward eight years. The firm needs to hire senior analysts and team leaders. The external talent pool is thin — every comparable firm made similar automation decisions. Internal candidates who joined post-automation have significant AI capability but limited commercial judgment. They have never experienced the formative years of working through a difficult due diligence process manually, making mistakes, being corrected, and developing the instincts that experience builds.
The firm has a skills gap that no AI tool can fill. Because the gap is not in technical capability — it’s in judgment. And judgment, unlike analytical skill, cannot be automated into someone’s career retrospectively.
This is what I mean by a leadership timebomb. The detonation is delayed. But the charge is already set.
What leading organisations are doing differently
The most thoughtful organisations I work with are not simply automating roles and moving on. They are actively redesigning career pathways to account for AI’s presence — and doing it now, while they still can.
They are creating hybrid roles. Rather than eliminating entry-level positions wholesale, they are redefining them. Junior staff still engage with complex work — but alongside AI systems rather than instead of them. They review AI output, identify errors, ask the questions the AI doesn’t know to ask, and develop judgment through exposure to real problems, even if the mechanical execution is handled by technology.
They are investing in accelerated rotational programmes. If AI compresses some of the time required to build technical competence, that creates an opportunity to accelerate breadth of experience. Smart organisations are using that freed capacity to give early-career professionals exposure to more business areas, more clients, and more contexts in a shorter time — building the commercial range that technical depth alone cannot provide.
They are building AI literacy into onboarding. The expectation is no longer that new hires will spend months learning to do the work before they encounter AI tools. They learn to work with AI from day one. But crucially, they learn to interrogate AI output, not just accept it. The skill being developed is critical judgment, not passive consumption.
They are redefining what “entry-level” means. The question is shifting from “can you do the work?” to “can you direct the AI that does the work, and do you know when its output needs human intervention?” That is a higher-order capability requirement than the previous baseline. Meeting it demands more deliberate investment in early-career development, not less.
A practical audit framework
If you lead an organisation that is deploying or planning to deploy AI in functions that currently involve significant entry-level work, here is a framework I recommend.
Map your roles against AI automation likelihood. For each role at meaningful risk of significant automation in the next three years, document not just what the role does, but what it teaches. What commercial skills, contextual knowledge, and professional habits does someone develop by doing this work for two years? This is the learning asset you need to preserve.
Identify the gaps your automation creates. If the work being automated was previously serving as a training ground, what is your replacement mechanism? This question needs an answer before the automation goes live, not after the capability gap becomes visible.
Design deliberate learning pathways. This means building structured experiences that provide the exposure, challenge, and feedback that entry-level roles traditionally provided — through new role types, rotational programmes, mentorship structures, and AI-assisted learning environments that require judgment, not just execution.
Feed findings into workforce planning. The output of this audit should directly inform your hiring strategy, your training investment, and your succession planning. The talent gap risk is not abstract — it can be quantified, planned for, and mitigated if you act now.
The strategic imperative
ANZ organisations operate in a relatively small talent market. We do not have the scale buffer that larger economies enjoy. A missing cohort of experienced mid-career professionals hits harder here than it does in the United States or the United Kingdom, where the talent pool is simply larger.
This means the urgency of building deliberate talent development strategies that account for AI’s impact is higher for us, not lower.
The good news is that the organisations that recognise this risk now have a genuine competitive advantage. While others are automating without thinking about what is being lost, they are building a talent pipeline that will be genuinely differentiated in five to ten years.
At Quanton, we believe that people should be empowered by technology, not replaced by it. That is not sentiment — it is strategy. The organisations that will thrive in the AI era are those that invest in human capability with the same rigour and intentionality they bring to technology investment.
The future workforce is not all human, and it is not all AI. It is a thoughtfully designed partnership between the two. Building that partnership starts with ensuring the humans in it have had the chance to develop the judgment, experience, and capability that makes them genuinely irreplaceable.
That development begins at the entry level. And the time to protect it is now.