By Alex Roberts, Director, AI in Action, Department of Finance
To deliver better outcomes for Australians we as public servants need to know what we are talking about. But what does that look like when the world around us keeps changing, and the answers keep shifting?
“We say we’re evidence-based, but our calendars say otherwise"
A flippant witticism (thanks ChatGPT!) but illustrating a telling point:
- The recent Australian Public Service (APS) Census indicated that for over half of us in the APS, we feel the workload is either slightly or well above capacity.
- Last year there were over 5 million journal articles published, just another signal of how much new information we are all grappling with.
Dedicated as we are however, we have to focus. And public service is about our judgment, not our omniscience. In a changing environment, our sense of professionalism needs to move from having all the answers, to contributing to a learning system. We can’t be across it all as individuals, but together we can improve by challenging our assumptions, testing our work, learning faster, and sharing what we learn.
And these are all behaviours and habits that AI can helps us with.
AI as critical thought partner
We all have assumptions and biases – but AI can make it easier to expose and test those. A simple exercise is to use AI to test a drafted proposal and to ask questions such as “What assumptions is this work built on that deserve questioning?”
Putting yourselves in the shoes of others
One of my favourite ways to use generative AI tools is to ask it to approach things from a different perspective to mine. AI allows us to outsource some of the cognitive load of thinking from different perspectives without sacrificing what matters—ensuring we hold our own thinking up to scrutiny. This can be as simple as asking, ‘Critique this from the perspective of x stakeholder group who might view y policy as an unwelcome intervention’.
This isn’t about trying to shortcut consultations or cheat ourselves of a full appreciation of the concerns of stakeholders—rather it is about helping have respect for their time and position by getting insights to help enrich those conversations.
Fleshing out our ideas
One of the hardest bits about putting forward an idea is that it can be clear in your head, but conveying that conception and the potential to someone else can be difficult. That is why design thinking places such emphasis on making things tangible prototyping helps us flesh out ideas, which makes it easier to see how (or if) they would work, and helps others provide a response about what they do or don’t like.
Asking an AI model such as Claude to help you mock up a working prototype is now easier than ever. By making ideas more ‘real’ quickly and cheaply, we can focus our efforts more on the important questions of whether it would work and would it be a good strategic fit, rather than investing time and money into developing out something that we’re not even sure about.
Iterating and learning faster
The IPAVentures team at IP Australia has been developing a new platform, IP First Response. The team’s strategic designer used AI prototypes to iterate features in between stakeholder and customer testing interviews, meaning that new insights could be tested in real time. By making it easier to respond to feedback and showing the results, it was easier to get richer insights faster, saving not only the team’s time, but also that of the stakeholders investing their time.
AI can make us better publics servants by making the right behaviours easier
Some colleagues already test assumptions, prototype at pace and shift easily between stakeholder viewpoints. For the rest of us, AI can be a great platform for pushing ourselves to better the best public servants we can be.
