TEDx talk January 25 2012 in Kortrijk
The talk in this TEDx is about the non-reductability of reality to our models and the unavoidability of emergence.
http://www.youtube.com/watch?list=PLE285E53633293B3F&feature=player_detailpage&v=mgPbR6tIA60
Transcription of the talk:
Recently we celebrate the sixth birthday of my oldest grandson. Sometimes I would stimulate his imagination saying for example “mmm, this is a tasty dragon's blood” when he drinks his tomato soup. Than he would look at me with twinkling eyes and say: “Va, (that is how he calls me), this isn't real, isn't it, you are joking, no?”. Children, at that age, are still so creative and open minded that they very soon need to make a distinction between reality and how reality gets real!
…
What should we answer to this kind of questions? Can we help them to learn that it is impossible to experience something arbitrary but meanwhile that something arbitrary will happen also? Can we define a clear criterion in order to make out whether one is dreaming or not, whether something isn't just happening in ones imagination, whether one is lying or joking?
My answer is clearly “yes we can”. I even will demonstrate that this is at the core of emergence, the subject of this conference.
Let us demonstrate this with capturing the reality of throwing a die.
...
Observe the amount of pips showing. Let's call this the first case.
Let us study the difference now by not throwing the die, but let us all imagine that we throw the die.
…
Observe the amount of pips showing. That is the second case.
…
How many people did observe a 6, …, a 3?
Isn't that strange? To give an answer to the amount of pips showing you have to make a decision in the second case that you cannot make in the first case. Apparently, when we do something real, something else is emerging that we could not choose, whereas, in our imagination, events can happen only when we deliberately choose for it. Conclusion: randomness and emergence only exist in reality, you cannot randomly think something because you are the originator of the thought and therefore randomness cannot occur within that realm.
But is this not a to hasty conclusion?
Let us model randomness using some today's technology. To model that I cannot and will not take a decision, I can use a random number generator. By calculating the value modulo 6, I can use the result as the amount of pips showing when throwing a die. Let's call this the third case. Thanks to my programming colleagues I show you now how I throw a die in this model.
…
The result is...
Let's us study the difference with the first case. As in the first case: I throw the die...
…
I observe the pips showing, but
…
I also observe that the die has one of its edges showing in a very particular direction. If I use the north as reference I can measure that direction precisely using a compass. Let's do this:
…
I find...
Isn't that strange? When I throw the die again, it becomes very obvious that I cannot choose in advance the direction which is however realized very precisely in reality. If you recall the first case than you come to the conclusion that: not only pips were showing in the first case, but simultaneously also a very precise direction that I could not choose. This means that this last case is not different from the first case. So, let's call this last case also the first case. Apparently, when we do something real, something is happening also that we could not choose. In our imagination this could only be modeled afterward, what we did with the model based on a random number generator.
But is this not a to hasty conclusion?
Let us model this again using some more sophisticated technology. To model that I cannot and will not take a decision, I can use a random generator for pips and one for direction. By calculating the value modulo 360, I can use the last result as the direction showing when throwing a die. Let's call this the forth case.
I show you now how I throw a die in 3D, again thanks to my colleagues.
…
The result is … amount of pips showing and in the direction of …
But isn't that strange? In what direction is the direction north? By creating the model I had to take an arbitrary decision, a decision that I cannot take in reality without causing vagueness in my communication: we all have to agree on the direction north. Thus in my model I have to indicate precisely what I call the reference north. Conclusion: in my model I have to take a decision that I cannot take in reality.
But is this not a to hasty conclusion?
For sure, it is not a problem of projection. I can always project the images on the same flat surface where I throw the die and measure the direction north as I did with my compass and use that direction as a reference. Let's try to model this.
But... wait a moment. When I throw the die again, it becomes very obvious that not only I cannot choose the amount of pips, or the direction of edges, but also I cannot choose a precise position relative to the environment. If you recall the first case than you come to the conclusion that not only pips were showing in the first case (what I asked you to observe), but simultaneously a very precise direction that I could not choose and simultaneously a certain position in the environment that I could not choose either. This means that this last case is not different from the first case: all these aspects happened in the first case also. So, let's call this last case again the first case. Conclusion: when we do something real, something is emerging or happening also that we could not choose. In our imagination events can happen only when we choose for it.
But is this not a to hasty conclusion?
Let us model this now using some still more advanced today's technology. To model that I cannot and will not take a decision, I can use a random number generator for pips and one for direction and one for the position relative to the environment. Let's do this, even how complicated this will become to model the environment...
…
And did your hear the die falling?
And how high the die was thrown?
Etc...
I hope everybody can agree on the conclusion that we have to stop somewhere. We have a first case and than there is no limit on our creativity to come up with models as a second, third, forth, fifth,... case. Apparently a lot of agents or stakeholders interacting with this kind of tests could come up with other and different observations.
Apparently, when we do something real, we have to let happen also something different that we could not choose and this depends greatly on the kind of stakeholder and the kind of context. If we would like to go in search for something absolute in reality, this is the kind of reality we are referring to. If we want to study this and to influence what happens in reality we need not only models, we need prototypes interacting with agents in their context so that also something unexpected can happen. We have to respect all kind of stakeholders and their creative input.
How is this related to the subject of the conference? Let's list some observations using a different language:
Apparently reality has always more aspects than what could be expected. There are always emergent aspects. In communities with different interacting agents more aspects are emerging and need coordination: an emergent collective is born. An emergent collective is NOT a model but can be modeled differently by all interacting agents.
To handle these aspects we need creativity. This means that we need a variety of models to be able to control. This is called the “requisite variety” that is introduced and studied in cybernetics and systems theory.
Apparently the basic feature of reality is that we choose for some ordered model but that also something different than that order will happen, which is being studied in chaos theory. The butterfly effect or the high sensitivity to initial conditions, where small differences in initial conditions yield widely diverging outcomes shows the limitations of all our models of reality, even when they are deterministic. But also: randomness for one aspect does not imply randomness for all aspects. Our creativity to come up with non-random models is not limited.
Let me conclude by saying that we need a new science, a new theory and prototyping applications based on the only distinction: something and something else. This precisely is being developed at this particular place and time in reality where all kind of interacting agents are being respected. Join my excitement for creativity but also for the unexpected aspects of reality that we let emerge or happen at this very place and this very moment in time.
Thank you.