10 analogies from AI x GOV

By GovInsider

Big ideas from our flagship summit in short sentences.

We’ve just wrapped up our virtual flagship summit, AI x GOV! Speakers from all over Asia Pacific and the world discussed topics ranging from AI-driven citizen services and edutech to healthcare and cyber diplomacy.

Here are 10 big ideas on the future of public service innovation condensed into neat analogies.

1. Regional cyber cooperation is about guarding your neighbourhood

“If you see your neighbour’s house burning, do not point and laugh because yours will be next.” Gaurav Keerthi, Deputy Chief Executive (Development) at Singapore’s Cyber Security Agency shared at the Cyber diplomacy for a safer future panel.

This isn’t just a matter of goodwill. Countries have to help each other “because that fire spreads downwards”, he added.

ASEAN understands this, and member countries work together to share information every time an attack happens. “We immediately pick up the phone and say, ‘Are you seeing this too? This is what I have, what do you have?’,” he said.

The region’s approach to cyber cooperation could be a model for others. Countries have made individual commitments to the UN’s 11 cyber norms for responsible state behaviour. But ASEAN is the only group to have done so at a regional level, said Johanna Weaver, Director, Tech Policy Design Centre, Australian National University, at the same panel.

2. Defending healthcare systems is like protecting a child

"It takes a community to care for a child. It also takes a community to protect that child,” said Leon Chang, Assistant Chief Executive of Integrated Health Information Systems (IHiS), Singapore, at the panel on Protecting healthcare beyond its four walls.

It takes everyone in healthcare to protect healthcare, he added. This means that every staff interacting with the healthcare system from IT teams to clinicians, and even third party vendors, has a part to play.

For example, every staff member needs to be alert against phishing attempts; and everyone has a different role during simulation exercises of cyber attacks.

3. Cyber regulation is not a question of the pathway, but the destination

When sensitive data was kept on paper, governments didn’t mandate that it can only be kept in five storey buildings. In the digital age, it remains the case that regulating “the pathway” to cybersecurity is ineffective, HE Kersti Kaljulaid, President of Estonia 2016-2021, pointed out in her keynote address.

This was an analogy she used to explain that governments should focus their attention on securing data. As long as organisations can demonstrate that their data is protected, cyber regulation does not have to prescribe how every organisation should secure itself, she said.

4. A classroom of students is a classroom of unique paths

“A classroom of 20 is a classroom of 20 different paths,” shared Anna Korpi, Counsellor, Education and Science, Embassy of Finland in Singapore at the A classroom of one: Personalising the student experience with AI panel. Finland adopted a “deep personalisation” strategy to support the uniqueness of every student, she highlighted.

New tech tools are supporting Finland’s teachers in this, especially with time being a “scarce resource” during the pandemic, Korpi said. One tool helps to collect data on student wellbeing, rather than teachers having to gather that information manually.

This gives teachers more time to interact with and supports students facing challenges, she explained. It could even be used to detect issues such as bullying.

5. Moving to the cloud is like moving house and getting to upgrade appliances

Just as shifting houses is the best time to upgrade appliances, moving to the cloud gives organisations the opportunity to replace unwieldy processes and outdated systems, said Lim May-Ann, Director of the Fairtech Institute and Emeritus Director at the Asia Cloud Computing Association.

“When you rearchitect your government services, you start to upgrade,” she explained at The great migration: Securing the move to cloud panel. If the fridge is getting a little old, you get to buy a new one. If a database is slowing an organisation down, it can explore alternative options.

The chance to review current processes and rethink what works best could birth more innovations, she said.

“Let’s buy a new fridge, let’s do something new.”

6. AI may become another tool in the toolbox for future public officials

AI algorithms currently take multiple teams and a bunch of different technologies to put together. But Yong Lu, Vice President, Shanghai Data Exchange, envisions a future in which AI will become a simple drag and drop tool for public servants.

“AI can just be like a tool in the toolbox,” he said at the Future of AI in government panel. Officials would be able to assign a task to AI and leave it to work.

There isn’t yet enough historical data to train algorithms to do this, he explained. But as time passes and the tech matures, AI will only get easier to build and use.

7. When it comes to managing data across government services, think of data as a pipeline rather than a pile

Data is not “a pile of stuff” waiting to be cleaned up, shared Li Hongyi, Director, Open Government Products, Singapore at the AI for inclusion panel. To manage good data, it may be useful to imagine a “pipeline”, journeying from creation all the way to collection.

For governments to implement AI-driven projects well, it starts with ensuring that all processes are digitised, rather than depending on manual processes. This may look like having data stored on central databases, rather than Excel sheets stored on “Simon’s laptop on the other side of the office.”

This makes it easier to maintain base standards across the entire data journey from the get-go. Dates can be standardised across all government services as DD/MM/YY rather than MM/DD/YYYY, for instance.

When it comes to providing good data, it has less to do with fancy data models and more to do with ensuring systems are digitised and representing data in standard ways, he shared.

8. Creating AI-driven citizen services is like building a mall

Governments wouldn’t expect a citizen to know how to build a shopping mall or the details of how one may affect traffic flows, noted Rushi Ramachandrappa, Project Lead for G20 Global Smart Cities Alliance, World Economic Forum at the Smart city convenience: simplifying citizen experiences with AI panel.

Nevertheless, they may still include citizens in defining the vision for what role a shopping mall may play in the community, he said. At the local level, it is common to include citizens as part of the ideation stage for such public projects, he explained.

Similarly, citizens can still work with technologists to co-create AI-driven citizen services that serve the community well even if they don’t understand the nitty-gritty of building such services, he highlighted.

9. In education, data doesn’t mean magic

When Ibrahim Bashir finished his master’s degree in Computer Science, he looked to create his own educational technology startup with AI. “With all the hubris of a tech guy”, he thought that getting his hands on data would instantly lead to magic in the classroom, he shared at the A classroom of one: Personalising the student experience with AI panel.

But he discovered that isn’t enough. In his work as Technical Project Manager in Educate Ventures Research, he realised that tools which prioritise students and teachers, rather than the technology, will make the most impact.

AI is “a transformational” technology, but that results in organisations trying to “brute force it” onto different problems, he explained. The best tools for education are ones that focus on students and truly examine the best ways for them to learn, Bashir said.

This is the study of a student’s “metacognition”, which refers to how students learn how to learn, he shares. Personalisation without understanding how each student learns is simply profiling, Bashir continued.

10. Not controlling access is like selling air tickets without allocated seats

When employees have too much access to the organisation’s network, it’s like getting a ticket to an airplane without a seat allocation. Passengers can sit “wherever they want” once they enter, illustrated Scott Hesford, Director, Solutions Engineering, Asia Pacific at BeyondTrust.

In contrast, having a least privileged access management model ensures that employees can only access precisely what they require, he said at the panel on Securing the digital ID. This is similar to a ticket that has a clear seat number for each passenger, he explained.

This helps government employees build trust in citizen services and ensure that data remains protected from ransomware attacks and other data breaches, he noted.