„We need a new definition of reality“
Professor Floridi, why are people are so scared of the future these days?
It's a normal reaction to uncertainty. You enter a dark room and the first thing you think is: Is there something hiding there? It would be irrational not to be scared by something you don't know. The second reason is that we are more and more surrounded by technologies we don't understand.
But there were technologies in the past that we didn't understand. Take the telephone: It was believed that it would be a socially disruptive force, breaking down the barriers between public and private space. What is different now?
The difference is that the new technologies we are developing are highly autonomous. Put together uncertainty and autonomy and it becomes reasonable to be a bit concerned about what will happen, e.g. if the thermostat heats the house while I am away. It means I will get a big bill. Should I be worried about this? Sometimes, yes. When uncertainty and autonomy come together in nuclear power plants, being worried is a good thing. But I don't think we get the balance right. We are swinging between the two extremes, a simple-minded Californian optimism – "everything is gonna be great!" – and the doom scenario – "everything is getting from bad to worse." The boring truth is in the middle, of course.
The possibility of machine intelligence is talked about as the ultimate form of human progress as well as a threat to our existence. Stephen Hawking suggested we should escape into space before AI takes over. Are such fears justified?
Are AI a danger for the job-market? Yes. Are they a danger in terms of who controls what? No. To some extent the whole debate about AI taking over is, in the best case, naïve, and in the worst case it's a suspicious way of taking the responsibility away from the humans behind the machines. Why not talk about the real problems?
What would the real issues be?
We should focus on the implications on human life. Not in terms of superintelligence taking over, that's science fiction. One of the main reasons the technology works is because the world is adapting to it, is becoming IT-friendly. We should remember that the technology we are building becomes more and more successful because we are adapting the environment towards it. Here is an example: We have a robot that cleans our home, Roomba, it works amazingly, but it doesn't go under the sofa. So we are thinking about buying a new sofa. It's us adapting to Roomba. Next thing, the house has corners, but Roomba can't go in corners. Do we buy a new house? The fact that we are adapting environments to machines is one of the things we should be considering.
Yet, you argue in your book, The Fourth Revolution, that the way we think about the future is still too anthropocentric considering that machines have ‘outsmarted’ us in a variety of ways.
Yes, it's staggering how anthropocentric our current view is. If we see a machine achieving a task, then we think in terms of our intelligence. If we consider ethical actions, then we put our only interests at the centre of the debate, not the whole environment. And we can see where this leads: We are, literally, on the brink of making this planet uninhabitable. Maybe it's time for a change.
But isn't the human being in a special position compared to everything else?
The human being deserves attention for at least two reasons. First, it's a glitch. We are, so to speak, an hapax legomenon, which is a word that only appears once in a book. The classic example is the word 'gopher' in the Bible, a special wood with which the arc is build. It's being used only once. An hapax legomenon is totally natural, but it's a glitch. It is normal, but also exceptional. That’s how I see humanity. We are part of being, but also the hapax legomenon of being. The other reason for humanity’s special position is that in the whole chain of being, humanity is the only moment where being becomes conscious, free, able to care and make a difference. We are the ultimate anti-entropic creatures. We better take care of this, because it's the only chance the universe has of things getting better. So yes, it’s because of our special position within being that we can expect more attention. But that's different from imagining ourselves as the only element in the ethical discourse. We are special in the same sense in which a gardener is special: as a carer and steward of reality. Not in the sense in which we are special when it is our birthday and our party. We can conceive ourselves still different from and yet as part of the whole environment. I call this the ontic trust.
Is there anything about the digital revolution that could facilitate such a change in self-perception?
There is the chance of finding a unified language. There is a way of looking at cyber-culture as unifying man and machine. That is a bit too much for my taste, but I think there is something in it. We can build a vocabulary that shows that we are more part of the whole than we ever thought. When we show that computers can do some things we thought only we can do, like play chess, this is humbling in a good sense, it is re-grounding ourselves in reality.
So the advent of autonomous technologies does mark a turning point.
Yes. I wouldn't be surprised if you fast forward a hundred years, look back and say: There really was a rupture in continuity. There is much more uniformity between the 1960s, 70s, 80s and 90s, than between the 80s and 2020. We passed a threshold, but breaking a continuity will require new thinking.
What could this new thinking look like?
If you step back sufficiently from culture you start seeing big patterns. Modernity is highly dependent on concepts such as things, entities, mechanisms. Even people who have tried hard to move away from this perspective, say French structuralists, still talk about and are completely obsessed with things. This i-Phone, this chair, that table. They can look at the world from the perspective of things, but it's still a world made of objects. Fast forward and we realise that the philosophy we live by is more about networks rather than mechanisms, processes and relations rather than things and qualities. It is time philosophy stops being a minority report in the history of substance.
What is special about a network?
In a network the interactions and relations come first, then come the nodes. This table becomes just the interaction of physical forces, the physical forces come first, and then there is the table. The individuals, you and I, are part of social networks. There is a total shift we live by. More and more people today think in terms of correlations, probabilities rather than mechanistic causes and effects.
Networks also tend to be thought of as non-physical. Is physical space being made redundant?
Yes and no. Processes and relations are as physical as objects. It's just that instead of talking about roundabouts, we talk about the roads that form the roundabouts. It's not digital versus analogue, people insist on this mistakenly. The truth is that there is no clear divide between the real and analogue on one side, and the virtual and digital on the other. What is real cuts across the digital and the analogue, and what is virtual as well as unreal does the same. A genuine friendship can be more real online than a false one in the “real” world; a cyber-attack or a digital robbery is as real as anything can be, but we can play and pretend to impersonate characters any time in our life.
We need a new definition of what is real, a post-Newtonian one, and I have suggested interactability as a criterion. If you can interact with it, it is real within that context, independently of whether it is an icon on a screen, or a pebble in the street. Ghosts and angels are not real because you cannot interact with them. Databases and monsters on the screen are real because you can.
But the body, for instance, plays a very different role in relationships online.
Yes, we have this form of disembodiment. It's similar to the shift in emphasis from ownership to usage. None of the computers in this room are mine. If the embodied object is there to bear the properties of a service – like checking emails –, then it doesn't matter who owns the object. We care less about what the actual physical implementation is, as long as the function remains the same.
Does this kind of thinking also apply to people?
The more we spend time online, the more this will happen. I am not saying that this is fantastic. It bears huge risks. But the more we have a disembodied engagement with each other, the more we can play with our own embodiment. That, in the end, seems to me to be a good thing. None of us has really chosen the body we have. Now, once you start having a body that doesn't work the way you want it to, then this disembodiment starts to be a welcome thing.
Another way in which technology seems to be adapting not only the world around us, but humans, is related to memory. On the one hand, digitalisation and cheap online storage make it easier to remember than to forget. On the other hand, as you write in your book, 'save this document' is not equal to remembering but to replacing its old version. With increasing digitalisation, will we still be able to remember and forget things the way we used to?
There is now a constant process of rewriting the past. There are archival initiatives trying to take care of this, but there are no real policies on what we should erase. Now, we should put this into perspective. We don't have much material about say local elections in 19th century Britain. The past was pretty stingy in terms of information. Today we have much more, but it is more fragile than past memories, as it's so easily rewritten. We live in a constant present, and that is the problem with the right to be forgotten. At the same time, this constant present is very fragile: it is rewritten and re-edited all the time – think of old web pages –, it can be destroyed, e.g. by malware, it can be erased – some new pictures come in, some old ones have to go because there is not enough space – and it can become inaccessible – imagine old floppy disks or files written with obsolete software. The fragility of a constant, expanding present is our new predicament.
Does this sense of editability lead to a smaller time horizon when thinking about one's future?
Yes, if you can change direction along the way, you don't plan for the journey, or less so. Reversibility and editability of your life mean that you don't have to do things too seriously. The analogy here is a typewriter versus a word processor. If you type and type wrongly, it's a hell of a problem. You get the line wrong, you need to throw hours away. With a computer, you drop a few lines, get it wrong, try it again. It's a different way of writing. And likewise with life. We may be tempted to live in a constant beta version mode.
This appears to have implications for political processes as well.
Politically, this editability is a problem. Today, we have a major problem with the environment, which is not reversible. You cross a time line and after that you lose the reversibility. Imagine the gulf current stops – Britain may become uninhabitable. And it would be irreversible. There would be nothing we can do technologically, so we would just move elsewhere.
This tipping point is something agents in a free market may fail to take into account.
When economists say that markets can take care of any problem as long as you have the incentives and disincentives in place, I tend to agree theoretically. But I disagree when they have to add a "sooner or later" clause. Because that temporal clause is what matters today, it is the reversibility clause. As far as the environment is concerned, we don't have that reversibility any more. Once you reach a point of no return, I don't care that 'theoretically', 'in principle' a market 'could have' – it didn't, and now we are in deep trouble!
So the perception of editability and reversibility brings with it a form of short-sightedness.
Yes, and some lighter sense of responsibility, an underestimation of the danger of reaching points of no-return. Every child, every woman, every man that dies in the Mediterranean Sea, that's a turning point you cannot reverse.
We have talked about the past, and about the future. Does the way we look at the past shape how we conceive of the future?
There is a beautiful essay by Nietzsche, "On the Use and Abuse of History for Life". He makes this point about how history constraints our way of thinking about the future, our ability to see things anew. Every time we move forward, there is less room for thinking freely. There are more constraints. It becomes more difficult not to be aware of all that has been written, has been said, has been tried. If you are a musician today, wouldn't it be so much better not to know? We are getting old as humanity.
Is our species ageing linearly?
Yes, this starts with modernity. That's the moment when we have the first Herculean effort to regurgitate the whole past in view of a new story. It's like with an individual: Every time something important happens, you are trying to make sense of all your life in view of that new episode. How often can you do that? The reason we can't fall in love too many times is that it takes such an impossible semantic effort to make sure that: 'No, the person I met was not the right person, it all happened so that I would meet you.' We can do that maybe twice, three times. With history it's the same, and I don't think that the accumulation of data has changed that.