Keynote by Janusz Cieszyński – Edge Institute Digital Governance Summit

Hello! My name is Janusz, I come from Poland. I have a question, for the beginning: how many of you here are from the public sector? So you want to do digital stuff in public sector? Yes. Cool.

So, how I got here was a very strange situation. I was asked by accident to give a few thoughts on what we did in Poland to push our e-government initiatives, and at the end it turned out that the most used word during that conversation was leadership.

So this is for the few people sitting in the first row. It’s mostly for you.

I was in the Polish government for about 5 years. Half the time I spent as the digital minister, second half I was in Ministry of Health. So I did electronic prescriptions and EID. What President Ilves was talking about, I did, only 15 years later, but you know, it’s still good.

So where are we, at least from my point of view? A lot of this inspiration for this slide came from a paper by President Ilves’s son. That’s true. Luukas, a friend, he wrote a paper, co-authored a paper on the agentic state. Why is this important? Because if you want to do anything in tech, government, or not government, AI is the word you should be following, and it really changes the way things look. Because everything that I wrote on the left was true even five years ago.

I mean, governments lag behind the private sector in terms of digital experience. I mean, you will not be judged in government against, for instance – on roads – when you talk about roads, you’ll be judged in the government against your predecessors, maybe some different governments nearby country. No, no. In digital, you will be judged against the ChatGPTs, Facebooks, Xs, whatever else. The apps that have billions of dollars in Silicon Valley being spent on. So you have to remember that.

And this creates a vicious cycle because if the government lags behind, then it’s super hard to attract the best talent. And I can tell you that even in Poland, we know that Romania is full of IT Talent. We know products from Romania, we know very talented Romanian engineers – so the foundation- you have. And I also talked with a colleague I met here, and what he said is, people in Romania are super smart about doing things with scarce resources. So you have something which is super important in the public sector, but the problem is that you are facing a completely different reality.

Because what AI did was it gave almost everyone the possibility to complete super complex tasks in very short time. Quality? Well, 6-7, but, it’s always like that. I mean, maybe you only work with super talented and smart people, but on average we work with average people, so results also average.

Second thing: time to launch digital services. The benchmark, the metric, the target is one day, and this is what’s going to happen. This is what is going to be possible, because it will be very easy to do something completely different from how we know implementing IT in the public sector. You will actually be able to just prompt a policy.

Maybe not now, but in a very short time — shorter than we expect — you will be able to give the agent the policy that it should be implementing. And this is like the hyper-Pareto stuff. Ninety percent of the tasks will be obsolete. You will not need to do almost anything to complete them, and this will free a huge amount of resources: first, to develop all of this, and second, to focus on the complex tasks which IT systems cannot do yet.

This is also important because I have heard a lot that, “It is bad for civil servants because they are not IT professionals.” But the most important person is not the IT person. The most important person is the one who knows the product. And when you think of a person who has a vision of how things should work on a big scale — well, in one word, that’s a politician, in theory.

So this is why the agentic state can be a revolution. I strongly recommend you read the whole paper.

And to guide this revolution, there is one thing you need — you need a leader. You will be told in politics that there are many things you cannot do for various reasons. So when I was in government, I was telling all my people: “We are the government. We are here to govern. We are here to make decisions. We are here to make things happen. And it is also possible for all of you. You just have to push really hard to do it.”

And all of these people you see on the screen — they did not get to their positions, their fortunes, their fame by following rules or doing things by the rules they found when they arrived. They mostly got where they are because they said: “Hey, we’re not doing it this way, we’re doing it my way.”

And imagine the people at Twitter when Elon Musk bought it. Imagine what happened to the people who said, “Hey, we can’t do that.” They are definitely not in the company anymore.

And yes, you sometimes make mistakes. You should make them less often rather than more often — but you have to be willing to make mistakes. True leadership in politics means having this guy or woman above you who says: “You can make mistakes, and as long as I see that you made them with good intentions and you don’t make too many, I will protect you from being fired.”

So this is why you need very strong leaders doing the digital transformation. And to do that, you don’t hear those people talking about the buzzwords you hear at govtech conferences. You don’t hear them talking about interoperability. It’s my favorite word. I hate it. I hate it. Never use it; never do anything about interoperability.

The only quote I found about interoperability was from Zuckerberg in a letter to shareholders — so that’s for Wall Street people. And Steve Jobs used the word interoperability when he was promoting his product. You know which product? The iPod, where he said: if you buy the music, it’s going to be yours and you’ll be able to use it. So interoperability was his way to bash the heads of his competition to push his idea for the product.

So don’t talk about interoperability. Talk about the product, and why the product is most important.

A quick case study: when the war in Ukraine started, within three weeks we launched the possibility for every Ukrainian citizen to obtain a digital ID in every local government office. Three weeks.

Maybe it doesn’t speak very well about our counterintelligence, but no one told us, “Hey, you should prepare.” Well, too bad. We still did it, and we made it possible for every person who came from Ukraine to use public healthcare, public social security, to pay taxes in Poland, and to use all kinds of other government services.

And why I got involved in this: the first idea was, “Hey, let’s give them paper or plastic IDs.” That would have taken so much time, it would have cost so much money. But I knew I wanted to launch the e-ID for Polish citizens, so I said: “Hey, let’s try this with the Ukrainian people. I hope it’s going to work.” It did.

And because of that, we issued one million IDs within 30 days, while maintaining the issuing of normal plastic IDs for Poles at the same time. We did not add a lot of new resources. We just used what we had, and we proved to everyone that it can be very quick and it can solve a lot of problems.

And why we did that is because from DIIA — the Ukrainian government app — we made it possible to create the Polish government app.

Hey, it looks fairly similar, huh?

And the first attempt was my predecessor, who said: “Poles should have a digital ID. And if you don’t think it’s a good idea, well, you’re stupid, because it’s a great idea.” How did that go? Well, not very well. She was fired because she didn’t deliver. She delivered a lot of presentations. She made a lot of slides. But she did not deliver the actual stuff.

And the biggest disagreement was with the Ministry of Interior. And who do you think benefited the most from the Ukrainian project? Who was in charge of the two million refugees that came into Poland in two weeks, and who was completely not ready for this? Well, yes — the Ministry of Interior.

It came to a point where, in the end, the Ministry of Interior said: “This Ukrainian stuff — that’s our idea.” Yeah. Right.

This is why, in the end, we were able… (short clip) — when they’re voting on the e-ID legislation in Parliament, lots of people voting, and the end result is: almost everyone voted in favor. Only the pro-Russian group voted against. True story. Maybe you have that too.

What we made possible is that because the product is good and people like it, everyone voted for it. And we were also open with our very fierce political opponents. For instance, there was this guy who said: “Okay, we don’t know what you’re doing there. So why don’t you open source the code?” I said: “Okay, no problem.”

So this guy became Deputy Minister of Digital Affairs. And within six months of coming to office, he went to Parliament and said: “For security reasons, this code cannot be open sourced.” Well, it’s not open sourced.

They were afraid of it, because they knew that if they open sourced it, they would have to do one thing a lot of people really dislike doing…

They would have to work hard to make it good code — better than it is right now — because if no one sees it, you can do spaghetti code, whatever. Sometimes it breaks — well, too bad.

So this is why what I wanted to tell you is: delivery is king. You have to have the best product, you have to have it quickly, and you have to have it in a way where you will actually be able to compete with the best.

This is the Polish App Store. I did this last week. mObywatel is the third most downloaded app in the productivity category. Third. It’s even better than Grok. And how did we do this?

First of all, we built a strong team in-house. Second, we did a whole thing around quality and design, and about how people use it. When we were designing the app, we told people: “These are the mock-ups. You can go online, you can comment. You can say: we don’t like that, we like that, we want this, we want that.” They even voted on the logo. And that gives you a community of believers. It’s like those people who, when you say “Maybe Elon wasn’t right this time,” you get 5,000 comments saying, “No, he was!” You need to build this kind of community around what you’re doing in government as well.

And because expectations toward government are so low, if you do anything right, you will instantly find a lot of people who say: “Hey, I’m super surprised — it’s good! I’ll support this.” It really works. And also, it’s good if it really works.

We have half a million ratings, and the average is 4.8 in the App Store — probably twice as much in the Google Store. We also did a survey; you can see the satisfaction results, and it shows that people really like the app because it was made for them and with them. This is key if you want to deliver good solutions and be able to say: “Hey, we were wrong — let’s do it a different way.”

This is true leadership. Of course, as a leader you should not say you were wrong too many times, but you should be able to say: “That was my idea from the start. I was right all the time.” And use what the people gave you. It really works.

And what mObywatel made possible is that when we had the launch date — the whole system crashed on the first day. It went totally underwater. And our political opponents started attacking us, saying: “Hey, you messed up again.”

And the most surprising thing happened. The people stood up and said: “Hey, well, you guys never did anything like that.” I will not use the words they used. So this is why it is super important in digital projects to be a leader with strong backing from the people. And the people are the users. We actually have more than ten million active users right now, and it is one of the things that makes us very proud.

And now the new government says: the previous government was the worst government ever — that’s mine — but there’s one good thing. This one. They really say that.

So, key takeaways: what I have learned as the Digital Minister, as a politician, as a leader. First of all, the AI component. I said at the start that it is a must today. You have to introduce it, and it will probably make your solution better, because you will be able to do things that required fifty clicks or fifteen clicks by just talking to your phone. That’s how people use their phones today. You have to use it.

You also should do what President Ilves was speaking about — that there are probably a lot of things already done in the AI area for you. Use them. But it doesn’t mean you should use American stuff and not do Romanian stuff. No. But it is best to make the first steps using proven technology, and then find the Romanian way.

Find the Romanian way.

Invest in design — this is super important. Go on your government website and ask yourself one question: “Is it pretty?” No, it’s not. That’s bad. Because to rewrite the whole code is going to take a lot of time. So get some talented graphics person. I don’t know — you had a campaign recently. All those people who were doing all those posts online — they’re talented. Let them design your government portal. It costs a fraction of the code.

And when it looks pretty — like in life with humans — if something looks pretty, you will be able to forgive it more things. If it doesn’t work very well… well, but it’s pretty.

Launch quickly. And the thing you can launch the quickest is the design. You can do that very quickly, and then do the app.

I mean, someone will tell you: “Oh, an app — that’s going to be so complicated.” Well, the easiest way to do an app is to just make a simple app. The web thing behind the app will be the website you already have, with some modifications. Just make an app that is basically a browser with the ID inside, and that’s all.

And you can say: “Hey, we launched a government app. It’s great.”

Put all the apples in one basket. Do a super app. Do a super app, because traffic generates more traffic. In Poland, when we have a big concert at the national stadium, you know what crashes? The driver’s registry. Because people, when they go to the concert, open the app to show their ID and then say: “Ooh, all those cool services, I can check if I have any speeding tickets or whatever.” So they start DDoS-ing the driver’s registry and it fails.

And when the driver’s registry fails during a concert, or there’s something else happening, or there’s an advertisement for the app on TV — that’s when it happens. So instead of having all the different apps, put them in one place, and you’ll get a lot of traffic from people who came in to use other stuff.

Each ministry needs a CTO, and even more importantly, a CEO-style mindset. Every government minister should know one thing: if you are going to do something in 2025, it is going to have an IT component. So you have to support the person who is responsible for this IT component. Even if he’s not very social, even if he thinks he is always right — you should support him, because you will not get far without the tech.

And the last thing — this is the hardest, or one of the hardest. You will always hear the same thing: “But we can’t hire the people. It’s government. We can’t pay them good salaries.” Yes, of course. And then you go to IBM or some other big corporation, and they give you a huge bill for whatever they do. But for that, every government has money — no problem.

So move the money from one place to another. If you don’t have extra money to invest in digital, this is your role as leaders: to make executive decisions like this and say: “We will take the political fallout for hiring talented Romanian engineers to do our e-government solutions.”

We did that in Poland. And when the government changed, they fired the people at the top — but everyone else is still there, because they are not stupid enough to fire the people who are doing a great job.