"One day we will be able to build human-like artificial intelligence – and we will have to"

What will tomorrow's world look like? How can we make it better with intelligent solutions? In a new series of interviews, our Co-CEO Ana Campos discusses these questions with people who have the answers. The series kicks off with the Swiss AI pioneer Pascal Kaufmann. With the Mindfire Foundation, Pascal networks the brightest minds in the world – with one goal: to decode the principle of intelligence.


Ana Campos: You have been working in the artificial intelligence field for over 20 years. What conclusion do you draw from all those years of research and development: What is "intelligence"? And what is "artificial intelligence"?

We still only know surprisingly little about what intelligence is exactly. A few hundred years ago, people believed that the brain was made up of cogs and small pumps because the clock mechanism was in vogue at the time. These days, many people believe that the brain contains a computer that will suddenly become intelligent if you can just make it run fast enough. I doubt these analogies. A much more interesting concept, in my opinion, is that the brain is probably a superorganism that consists of billions of small parts and can obey simple rules. Like a colony of ants or a school of fish. One of the newest concepts proposes that we are a kind of holobiont. This means that we consist of 95% microorganisms and only 5% of our own cells that carry the genetic material of the parents. That puts everything into a new perspective. For me, intelligence arises from the interaction of various elementary modules. That is why, in my view, the Internet and a crowd of people are intelligent – or a large number of cells. Intelligence has to do with surviving in a changing world, responding to stimuli and fulfilling goals. However, we still understand qualitatively very little about the principles of intelligence.

You started out as a neuroscientist. What is the brain for humans? Or to put it another way: what is a human without a brain?

I believe that humans have at least three brains. The first brain is what we know and have between our ears. The second brain is our culture. For example, a place name sign on the road is also a type of externalised intelligence that we outsource to our surroundings. Just like the knowledge in books that we pass down through generations. The third brain are the genes. There are many animals that can already fly or swim at birth. So there is a lot of know-how and experience about the world in the genes. If you look at it like that, we are made up of many individual brains. We are not really aware of that.

There are numerous theories about consciousness: personally, I have a lot of respect for Julian Jaynes, who taught in the United States. At the end of the 20th century, he postulated that human consciousness has not existed for very long, perhaps only since 1500 BC. Before that, humans were a kind of automaton. You can then quite clearly explain why there are phenomena like religion or how it leads to building statues of gods. Or how the concept of an immortal soul and how the "I" concept could have developed. I can recommend Jaynes's book to anyone interested in consciousness.

Can man be "explained" in a completely deterministic way or is there anything else that goes beyond this?

I believe that we obey the laws of nature and that everything can actually be explained – and everything would be predictable, provided that you knew all the parameters. This concept is sometimes referred to as the Laplace's demon: what if someone knew the movement of all the particles in the universe? Would they then be able to predict the future?

Then why live at all if we are slaves to the laws of nature?

It's like saying: I'm playing in a football match where there are rules that I have to abide by and that is why I don't want to play. There are a lot of people who enjoy participating in a completely pre-defined computer game. There is a great deal of attraction in achieving as much as possible within this set of rules and playing a good game. That is probably why I like Renaissance paintings, which depict the world as a great world theatre in which everyone has their role. I consider it very important that each of us plays our role to the best of our knowledge and belief – even if we are probably just puppets and subject to the laws of nature. Some people see the strings, while others don't, but it doesn't change the fact that we're subject to the laws of nature.

Popular culture often tries to convince us that the world with artificial intelligence is doomed. How do you counter such dystopian ideas?

If you look at the media, you will find extreme views like that of an Elon Musk who thinks artificial intelligence is very dangerous, or that of a Mark Zuckerberg who is enthusiastic about it. It is interesting that such extreme views are often represented by people who do not deal directly with the matter and who just scratch the surface. Imagine if 200,000 years ago the people of the Stone Age had said: "Guys, the fire is dangerous, let's put it out again". But they didn't; they said: "Let's tame the fire and make it work for us!" I think that was a good decision. And we should take the same approach to artificial intelligence. We have to build artificial intelligence to solve the great challenges of our time. A single human brain is simply not able to fully grasp certain current problems and then develop solutions. New challenges that cannot be solved by traditional approaches for long require new solutions and new tools.

One of the Mindfire Foundation projects has the working title "Promethia". Prometheus gave fire to humanity. Do you see yourself as a modern Prometheus? If so, why?

Prometheus stole the fire from Zeus, thereby questioning and changing the status quo. After Prometheus, the world of the gods was never the same again. According to Greek mythology, humanity was also created and inspired by Prometheus. Prometheus means "the forward thinking". I am intensely concerned with how our world could or should look in the future – and how we can solve the challenges that we face. In this respect, I sympathise with Prometheus. But I don't want to be a Prometheus chained to a rock over the millennia. I'll be careful about that.

pascal kaufmann_ana campos_klPascal Kaufmann, Roboy and Ana Campos at the headquarters of the Mindfire Foundation.

As a child, you wanted to invent a magic machine that would change the world. How close are you to this vision today?

I am getting a little closer to the magic machine every day because I have more and more components. In particular with the Starmind technology, which networks a large number of intelligent people worldwide into a kind of superorganism. I also have great expectations of virtual reality and augmented intelligence technologies.

You have already won a variety of awards including "Newcomer of the Year" and "Digital Entrepreneur of the Year". Who is the person behind the accolades?

I stand up for "Team Human", I like to discover new things and I don't just accept the status quo. When someone tells me something is impossible, it motivates me more than it scares me away. If America or the moon hadn't been discovered yet, I would be sitting in a boat or capsule right now. I want to push the boundaries and believe that one day we will be able to build human-like artificial intelligence – and we will have to.

What gets you angry and what keeps you grounded?

I don't really get upset. It takes a lot. However, I am frustrated by people who are hesitant and indecisive and who stall things that are actually predictable and inevitable. And what keeps me grounded in reality is how anxious people can be, especially when it comes to new things. And how often people stick to the status quo. In this context, it is also interesting how different the generations think. A young person has very different motives than an older person who is at the end of their professional career. At universities I often find a conflict between the generations, whereby young researchers sometimes work very differently and are motivated, while more traditional researchers often hold higher positions and like to defend the status quo or reject ideas from the outset. That needs to change.

What makes you happy?

I am happy when I achieve difficult goals or discover new things. When I see that as a team you were able to achieve something that an individual couldn't. I am a fan of exceptional talents that make a big difference. Of course I am always very happy when I get to spend time with my loved ones or eat pear sorbet.

What do you wish for the world in 2040?

I want a world in which there is human-like artificial intelligence. A world where we have decoded the principle of intelligence. And where we solve or have solved the challenges of our time using technology. Maybe we should organise two or three spare planets, because with 10 billion people it seems a bit of a tight squeeze. I also wish for a world where there is no shortage of resources and where everyone is doing well, a paradise on earth. Even if this vision is ambitious, we should at least strive for it.

With Mindfire you have access to a network of numerous talents and thereby answers to all kinds of questions. What does Pascal Kaufmann not know yet?

Although we can draw on many talents across the globe through Mindfire, the principle of the brain has not yet been decoded. I wish that the principle of intelligence is decoded soon. I would like to know how it works, how to build such an intelligent organ out of individual cells, which, in interaction with the environment and a body, produces such gigantic things.

 Thank you for the interesting chat, Pascal!

About Pascal Kaufmann:

Pascal Kaufmann (born in 1978) is the Founder and President of the Mindfire Foundation, which has set itself the goal of creating human-like artificial intelligence. He studied neurosciences at ETH Zurich and Northwestern University, IL, USA, and specialises in artificial intelligence. In 2010, Kaufmann co-founded Starmind, an internationally successful AI tech startup that offers self-learning know-how networks for companies. He was also involved in the construction of humanoid robots and neural networks at the artificial intelligence laboratory at the University of Zurich for several years.


Topics: Innovation Interview