The AI enthusiast:“AI should be approached strategically but right now I mostly see an opportunistic approach and fluff about ethics…”

In October it is time for the annual European Forum on AI & Data Ethics 2020 conference, which will be held online as well as onsite in Copenhagen. The program is filled with interesting cases and brilliant speakers. One of them is Dan Rose Johansen. He has a slightly different take on AI compared with many of the other speakers.

Dansk IT has had the pleasure of meeting Dan for a talk about his work and his expectations for the conference and about what he will share with the audience.

You are one of the speakers on the European Forum on AI & Data Ethics 2020 conference, what will you share with the audience? Please outline what you will be discussing during your presentation/debate with Daniel Safai from Datamaga?

My goal is to make AI understandable and practical from a business perspective. When I hear people talk about AI, they usually talk about very technical details, the very far future or about ethics. I want to talk about what AI can do right now and how you do it. AI should be approached strategically but right now I mostly see an opportunistic approach and fluff about ethics.

What are your expectations for the conference?

I expect and hope for some good debate and some good networking.

What do you hope your participation will contribute to?

I hope I will contribute with an alternative opinion on a lot of the issues discussed. I also hope to contribute with my own hands-on experiences from building an applied AI-company from the ground up. I have a background in both economics and as a programmer, and, hopefully, I can help people make the connection between technology and the business aspects.

The tagline of the conference is that there is “No AI without Data. No Data without Ethics”. What is your opinion on that?

It is a good tagline. However, I look forward to challenging the meaning behind it. In my opinion, not all ethical issues are related to the issues of using data or AI. Data and AI are just putting the issues in our faces and making us practise what we preach. It is making us uncomfortable.

In an article on your blog, you state that ethics are not relevant to AI. Please elaborate on why you think that is so.

My basic point is that new technologies should not be held to a golden standard. If new technologies keep status quo on ethics, then it is fine. Golden standards are a risk to new technologies, such as AI, that might bring important benefits in the long run.

At the same time, I can mention several examples of AI technologies in need of ethics like self-driving cars, race issues and killer robots. In my opinion, AI is not bringing any new ethical issues. The ethical issues are already there, and if it is really a problem, why not fix it now? If "who should the self-driving car kill in an accident" is such an important issue, then why do we not have the "who should the human driver kill in an accident" discussion now? So, let us get all these ethical issues out and not wait for the technology to rise. 

Why did you choose to specialize in AI?

That was a pure accident. I co-founded the company Paperflow, which set out to capture data from invoices, so small business owners did not have to. AI just happened to be the right technology for us to use. Over time, I became very passionate about AI, since it gives us almost endless possibilities with challenging problems.

Today, I help companies utilize the power of AI. I do that mainly by educating and working with how to strategically use AI to improve business results. Surprisingly, the technology is the easiest problem when working with AI. The human and strategic aspects are usually much where the problems are rooted. It is in that tension between technology and people that I like to be.

What is the best thing about your profession?

It is the combination between human, tech, and strategy. AI effect its environment most considerably. At the same time, there is no exact right answer to AI problems. Whatever decision you make in AI, it always comes with a trade-off and makes it exciting.   

EFAD 2020  

This concludes our interview with Dan.

Dan will set AI straight on Day 1 of the conference European Forum on AI & Data Ethics 2020. He is on for 3.30pm to 4pm.

Check out the program here.