Q&A: Richard Pope, author of Platformland
|
Richard Pope is a founding team member of the UK Government Digital Service, the first product manager for GOV.UK and author of Platformland – An anatomy of next-generation public services.
At our April 2025 #DigitalBC livestream, Richard and host Gen Lambert, Chief Digital Officer, discussed ideas from Platformland, defined connected services and covered the challenges people can face on their journey to implementing them successfully. If you missed the livestream, the full recording is available on YouTube.
Following the livestream, Richard answered some questions we didn’t get the chance to address in-session.
How do you think artificial intelligence (AI) should be integrated into government services and how can public trust be maintained?
Great question, because “integrated” is totally the right framing. AI will be part of services, not the service. I really like how British designer Matt Jones talks about distributed intelligence.
I fear the current fetishization of AI in the public sector makes that harder. Where I live in the UK, it sometimes seems that ministers want “an AI answer.”
There needs to be a much better understanding of the parts of problems that the various types of AI systems are good at, such as transcription, pattern recognition in images and predicting next steps in a process. Those can be turned into repeatable patterns.
Three levels of trust
- Trust is multifaceted – sometimes people talk about one type when they mean another. At the most basic level, people need to be able to know that they will get the right outcome to a high enough level of confidence.
- We also need digital trust: that data has not been tampered with, that I really am talking to a public official and not a random bot.
- There also needs to be collective public trust that a public institution is delivering on its mission (for example, providing healthcare), not something that appears contrary to that immediate public good.
All of this is hard in data-hungry, non-deterministic systems.
Teams developing systems which include AI in the mix need to be radically humble about sharing what is working and what isn’t. As I set out in my book, it also needs to become a matter of public record how digital systems work and how they are changing.
Which skills will be most important for public service workers as digital services grow over the next decade?
The short answer is: delivery. There’s no substitute for having done the work. Senior people need to have that hands-on experience on their CVs.
I think of value in the development of digital public services as a mix of capacity to deliver, meeting user needs and political intent. The public sector needs more people who are empowered to make the often messy decisions about trade-off between those competing priorities.
What should we consider when deciding how to connect different services in a “Connected Services” initiative?
Start by measuring administrative burden and don’t be constrained by the boundaries of the current service. For example, a social security service might create burdens in the courts system.
Before thinking about the design of a new end-to-end service, look for opportunities for “loose joins” between services. These sort of “composite services” are particularly important in areas where one user’s journey through the system might be very different from others’.
How important is creating and supporting a learning network with small test-and-learn innovation teams across organizations? And why is that?
There are some problem spaces which are essentially about usability.
“Can a user understand this application process and get the correct outcome?”
Out of the three measures of value (capacity to deliver, meeting user needs and political intent), these are mostly about meeting well understood user needs.
But there are some classes of problem (I’d argue the majority in the public sector) where you only learn what the right thing to do, what the user needs are and if you can meet the political intent, by creating a relationship with a small number of real users in a live service and then testing what works. This is more about building capacity to deliver.
What funding models could support a product mindset in digital-first government services?
Government needs funding models that mirror the incremental nature of digital delivery. It also needs to be able to scale funding for the long term, especially for digital and infrastructure where the value is created over a longer time period.
You can learn more about Richard on his website, RichardPope.org. Make sure to catch our future livestreams by signing up for the Exchange Newsletter.