Daniel Sunde-Hansen in front of a screen showing his own agent system

Digi.no August 8, 2026 AI Agents Digital sovereignty

Daniel has brought AI agents to life

"I can simulate processes overnight that would otherwise have taken months."

Daniel Sunde-Hansen is a PhD candidate in leadership at BI Norwegian Business School and brand new in the job as Partner for AI Strategy at Skyetec. Everyday work runs noticeably smoother with new AI agents on the team, and every digital colleague knows its place and its role. At night they argue about his doctorate. During the day the assistant is opened with an NFC chip implanted in his hand.

Originally published in Digi.no on August 8, 2026. By Axel Nyborg and Mia Huse. Photos: Mia Huse. English version for dansunhan.com.

The Innovator and the CFO

The AI agent "the Innovator" has few inhibitions and keeps producing bold proposals.

"It often challenges the very premise of the problem," Sunde-Hansen says.

The troublemaker will happily claim that a process is rigged wrong, and that he is trying to improve something that should not exist at all. "The CFO" is of an entirely different temperament.

"This AI agent consistently sees the challenges inside the opportunities, such as the proposals becoming too expensive or too demanding."

The value sits in the collision between them.

"The discussion can run all night, through hundreds of iterations. You could never have had humans sitting and debating for that long."

Some things he still refuses to delegate: what is worth investigating, which paths to take, and what the results actually mean.

"That is where human judgment becomes more important, not less."

The screen shows the agent system Sunde-Hansen has built, with tasks and cards organized as a map — The system is homemade from the ground up. Daniel Sunde-Hansen is the one making the final decisions. Photo: Mia Huse / Digi.no

Digital bickering

Sunde-Hansen programs layers on top of the agents that force Claude, ChatGPT and other local models to talk to each other.

"The platforms do not really like it, so it takes quite a bit of work. They learn from the interplay between the agents and gradually get better at coordinating them."

When the agents are forced to challenge each other's reasoning, they turn into merciless colleagues who refuse to accept easy solutions. It is precisely in this digital bickering that the magic happens, according to Sunde-Hansen.

Digital bickering in practice: the agents challenge each other's reasoning. Video: Daniel Sunde-Hansen

He has cloned his supervisor

He has also developed custom AI agents that simulate the academic perspectives of various BI professors. Through the night they discuss his drafts and hand them back covered in red comments.

He has built avatars of himself and of his supervisor, Jan Ketil Arnulf, among others. Every article Arnulf has published has been fed into an AI agent, so that it can reason from the perspectives and arguments in his supervisor's work.

"The virtual Professor Jan Ketil Arnulf knows everything the human Jan Ketil Arnulf has written before."

"He finds it very funny. He researches artificially intelligent leadership himself and has encouraged me to experiment," Sunde-Hansen says.

They speak and gesture

"It is not just for fun that I make them look like humans, gesture and speak. It makes the dialogue more natural and makes it easier to explore different perspectives."

Not everyone is equally thrilled about getting an AI twin. So far, four colleagues have given him permission.

"Some find it more problematic than others, and they may think it seems a little threatening."

The CFO speaks and gestures in conversation with Sunde-Hansen. Video: Daniel Sunde-Hansen

The assistant in his hand

The assistant is opened with a small NFC chip implanted in his hand. When he holds his phone against the chip, he gets access to the system. Hold a phone up to it, and his heart rhythm, his website and his LinkedIn profile appear.

"It has access to my calendar, my contacts, my email and my health data," he says.

"If I need to check something on my phone during a meeting, I can pass the phone over the chip and bring up exactly what I want to look at."

Phone against the chip in his hand, and the assistant opens. Video: Daniel Sunde-Hansen

He has also built a mechanism that generates music based on his heart rhythm every morning.

"I get calm jazz based on my pulse, for example," he says.

Sunde-Hansen points to the NFC chip implanted in his hand — The NFC chip in his hand opens the assistant. Photo: Mia Huse / Digi.no

Preserving autonomy

For Sunde-Hansen it is essential that the solution he uses is his own.

"ChatGPT recently asked whether I wanted to share my health data with it. I said no."

He says he would never share all the information he shares with his AI assistant with an American or Chinese agent.

"Because I built the system myself, I know exactly which data goes in and out."

"To preserve autonomy, both as individuals and as a nation, I think it is important to develop our own layers on top of the large systems."

"Should be replaced by machines"

Has AI taken over your life?

"Well. When I go for a walk alone, I often talk with my agents. I can hear how their work is going, give them new tasks, or ask them to explore an idea further."

Can there be negative consequences of this kind of AI use?

"Anything can be taken too far. AI can be used to exploit human weaknesses. My answer to that is not less AI, but better AI. We have to develop systems that serve the interests of users and strengthen people's ability to make good choices."

Daniel Sunde-Hansen in conversation during the interview — Daniel Sunde-Hansen believes you have to take part in the future in order to understand it. Photo: Mia Huse / Digi.no

The people who made his chip are working on a new version.

Are you getting it?

"I already have an agreement with the company to be first in line when the next version arrives," he smiles.

Sunde-Hansen believes much of the technology we see as exotic and alien now will be taken for granted a few years from now.

"I am not saying everyone should have a chip in their hand, but there are many possibilities."

He believes AI technology is here to stay.

"The people who are replaced by AI first are the people who are most like robots, paradoxically enough. Jobs where humans work like machines both will and should be replaced by machines."