AI is changing how machinery suppliers work and what customers expect. At APPMA’s Sydney Member Dinner, Dr Sean Gallagher urged businesses to build the human judgement and shared knowledge needed to direct the technology and remain accountable.
The late-night service call, the elusive fault and the quote that lost money have long helped machinery industry professionals develop their judgement. As AI takes on more routine work, how will the next generation gain that experience?
That question underpinned a presentation by Dr Sean Gallagher, founder of Humanova and a workforce researcher specialising in the future of work and organisational AI capabilities, at APPMA’s Sydney Member Dinner. Held on 12 October at The Grounds Factory in Eveleigh, the event brought together more than 60 attendees from the packaging and processing machinery industry.
Gallagher argued that businesses need to develop their people’s ability to make consequential decisions alongside their use of AI. His central message: “judgement is made, not bought”.
Changing customer expectations
For machinery suppliers, the pressure to adapt is already coming from customers. Gallagher described maintenance teams diagnosing faults before contacting suppliers, engineers preparing specifications with AI, and buyers expecting faster quotes that they can check against competing offers.
An attendee reinforced the point, describing customers using AI to identify suppliers, compare OEMs and arrive with specific questions about whether equipment could address their production problems.
Gallagher said AI’s value extends beyond saving time on existing tasks. Faster responses can reduce the opportunity cost of delays, allowing businesses to act while a customer’s interest remains strong. As routine diagnosis and quoting become more accessible, suppliers will need to recognise the value of their experienced people’s judgement.
Open-ended decisions still require someone to weigh competing priorities: what a customer actually needs, how to balance cost and uptime with food safety and sustainability, or whether to take on a job at all.
Before allowing AI to act independently, Gallagher proposed a simple test: “If it’s wrong, would we know? If it’s wrong, who pays? And can we undo it?”
Where the answers raise concerns, a person needs to own the decision and its consequences.
Building capability together
Gallagher encouraged businesses to move beyond isolated experimentation by individuals. Clear policies should protect customer drawings, pricing, intellectual property and personal information, while giving staff permission to explore approved business tools and share what works.
Humanova’s Work Value Pyramid starts with administrative and routine tasks, where errors are easier to identify and correct. People can build confidence before applying AI to more complex work, developing what Gallagher calls “AI intuition” – knowing when to trust a response, challenge it or change direction.
For staff starting out, he suggested changing the way they prompt AI. Asking it to pose questions before answering helps establish the customer’s needs and constraints, so the response draws on the context of the job rather than relying on assumptions.
He also urged businesses to protect the learning opportunities that automation could remove. Junior staff could complete the first pass on a low-stakes diagnosis or standard quote, making their own assessment before involving AI or seeking senior review.
Experienced staff should be encouraged to explain why they made a decision, recording the context, assumptions and trade-offs so others can learn from it.
Reflecting on how expertise develops, Gallagher said: “You don’t download that. You build it one consequence at a time.”
He encouraged businesses to assess whether people are getting better at working with AI, rather than measuring adoption simply by how often they use it. That shifts the focus to the capability teams will need as more routine work is automated.
