There’s an old Aesop fable about a farmer with a goose that laid a single golden egg each day. Impatient for more, he killed the goose to get all the gold at once. He found nothing. In trying to extract maximum value, he destroyed the very thing producing it.
Right now, in boardrooms across Australia, a version of the same conversation is unfolding: How do we get more out of AI? More use cases, more automation, more outputs, faster. It’s an entirely rational response to a 60-year productivity low and a technology that genuinely promises to help.
There’s no doubt that generative AI (GenAI) is enabling us to dramatically increase the volume of output. Within seconds, GenAI tools can perform analysis, generate polished slide decks, and write glossy, executive-ready reports. Tasks that once required significant time, skill, and cognitive effort are now accessible to almost anyone with a capable prompt and a decent model.
At first glance this looks like the saviour to our productivity woes. But look closer, and the early evidence is telling a different story. A 2025 MIT Media Lab report found that despite pouring $30–40 billion into GenAI, 95% of organisations studied were getting zero return.1
But the real story is not in the balance sheets. It’s in what’s happening to the people actually using these tools every day. When we look more closely at how GenAI is shaping human performance, an uncomfortable truth emerges. Using GenAI may not be accelerating your best people. In some cases, it may be diminishing the very capabilities that make them effective.
It turns out the golden goose is not GenAI itself. It is the people using it. And your best people may be the ones most at risk.
If you’re a high performer, AI may be making you worse
A recent study2 of business school students offers insight into why organisations may be struggling to capitalise on GenAI technologies.
The researchers asked students to use Gen AI tools while completing complex, time pressured tasks, and found something striking:
Low performing students improved their performance using GenAI. In contrast, high performing students got worse.
When both lower and higher performers move towards an average when they collaborate with AI, GenAI begins to look less like a productivity amplifier and more like a productivity equaliser. For your lowest performers, that may be welcome news. For the people your organisation depends on most, it’s a serious problem.
Why is this the case? The researchers argue that the underlying difference lies not in the technology itself, but in how people use it. AI is a tool and like any tool, value is driven by the mind of the person wielding it.
What’s happening in your brain
The researchers traced this back to cognitive load – the amount of mental effort being used by working memory at any given time.
Both groups heavily deferred to GenAI outputs. But what happened in their brains differed significantly.
Lower-performing students offloaded their thinking to AI and freed up mental capacity. Higher-performing students, by contrast, saw their cognitive load rise. Their mental effort shifted from composing to monitoring, evaluating, and integrating AI outputs. GenAI created more options, more iterations, and more signals to interpret. The result was rising stress, eroded clarity, and a constant tension between their own judgement and the machine’s.
This is the trap most organisations aren’t seeing. In the race to extract more from AI, they are quietly heaping pressure onto the very people who were already delivering the most. The goose isn’t being killed all at once. It’s being worn down, egg by egg, day by day.
Why this matters now in organisations
The implications go well beyond business schools.
First, your best people are at risk of burning out.
Research is finding that GenAI leads some employees to work at a faster pace, take on a broader scope of tasks, and extend into longer hours, often without being asked3. Over time, this expands the overall volume of work and the cognitive load required to find the signal in the noise. The next iteration of this – deploying swarms of AI agents to run tasks in parallel – risks intensifying the problem further.
Second, your teams are producing “workslop”4.
AI is so good at generating outputs that look polished and articulate but are actually vapid, incomplete, and unhelpful that someone had to invent a new word for it. The cognitive effort your team saves by producing “workslop” with GenAI gets handballed directly to whoever has to decipher, interrogate, and rework what they’ve been given. In many organisations, these are the very people who can least afford the distraction.
Third, there’s a gap between what executives believe and what’s actually happening.
A survey of 5,000 white-collar workers across the United States, the United Kingdom and Canada found that while over three quarters of C-suite executives believe their employees can save over half a day per week using AI tools, only 11% of workers agree with them, and 40% said that AI tools save them no time at all.
What can look like higher productivity in the short term may, in practice, be an accumulation of noise, task-switching, and a slower drift towards exhaustion.
These are not only questions about individual judgement. They are questions about organisational design and leadership. If AI is amplifying volume while eroding clarity, senior leaders need to look beyond the tool itself and examine the environment in which it is being used.
- What kinds of culture are we creating around speed, volume, responsiveness, and visible output?
- What are we rewarding?
- What are we accidentally training people to outsource?
- Where are we confusing activity with value?
How leaders can respond
This isn’t an argument against AI. It’s an argument for using it deliberately to protect the real golden geese: your best people.
Discern where AI creates value, and where it dilutes judgement.
A 2023 BCG study5 found that GenAI enhanced performance for 90% of participants on creative ideation tasks, producing 40% higher output quality. But on complex business problem-solving – tasks requiring nuanced judgement and contextual reasoning – participants using AI performed 23% worse than those who didn’t use it at all. While GenAI capabilities continue to evolve, the principle remains the same: Leaders need to stay current and think critically about what AI is actually good at. The tool is only as useful as your clarity about when to pick it up.
Think first and prompt second.
The more readily you reach for AI before engaging your own thinking, the more you risk eroding the judgement that makes you effective. GenAI can be sycophantic and opaque; it can reinforce first instincts, enable confirmation bias, and produce answers without making its reasoning visible. When you are working at pace, you’re less likely to interrogate what it gives you, and that is precisely when errors and flawed decisions begin to creep in.
Protect your cognitive capacity like the asset it is.
In the rush to apply AI everywhere, many organisations are either neglecting or actively undermining one of their most valuable resources: human cognitive capacity. Design work with not only the benefits of AI in mind, but the quality of attention required to use it well. Resist the impulse to default to AI. Build in real pauses. Make space for deep work. Slow down enough to think clearly. Over time, that is what enables better speed, not just more activity.
Strengthen judgement by exercising it.
Continued overreliance on AI can erode both cognition and confidence over time. Judgement, inquiry, attunement, presence, purpose, and accountability are not fixed traits that remain strong by default. They are capabilities that need to be consciously developed and continuously exercised. Left unused, they atrophy. The leaders who will get the most from AI are not necessarily the ones who use it most. They are the ones with the discernment to know when not to.
The farmer killed the goose because he couldn’t wait. He had something genuinely valuable and, in his impatience, destroyed it.
The golden goose in your organisation is not AI. It’s the people using it.
For senior leaders, the challenge is not simply to accelerate adoption. It is to lead with enough wisdom and foresight to ensure AI serves human capability rather than substituting for it. The organisations that will create enduring value will not be those that deploy AI most aggressively, but those that consciously regenerate and elevate the human system around it: the people, the thinking, and the leadership conditions that allow sound judgement to flourish.
