5. Let’s talk about God
When I was an undergraduate student of Mathematics, I once attended a lunch lecture by a Christian man talking about Christianity. If I remember correctly, it was entitled “Where is God in a world full of suffering?”. My good friend and I attended partly for the free pizza but also for the opportunity to question this man’s beliefs for ourselves.
I had not found my way to the undergraduate Philosophy program yet, which is probably why this lecturer had to endure my questions about believing and metaphysics¹ instead. After a discussion in which we agreed that there would always be suffering and in which we disagreed on God’s place, my friend and I were quite satisfied with the lecture. This was not because all our questions had been answered in a way which left no room for discussion, it was quite open ended. It was because — as my friend put it — we were content in the conclusion that the Christian religion is one of multiple manifolds² according to which we can understand the world and we admired the man for sticking to it.
Looking back at this discussion now, I can see the futility of my argumentative position. The gap between a settled Christian man and an aspiring atheist is not one that is possible to bridge with words alone. For me the existence of God was there to be questioned, whereas for him it was merely the role that God was playing in the experience of suffering that was at stake. Now that I would not call myself an atheist anymore, the gap remains: He believes in the existence of God, I am one of those still looking for a God to believe in.
Many discussions can be had — and have been had inconclusively, for which apologies go out from us philosophers to all tax payers — about the nature of our world. In this I mean the world at a very fundamental level, on a metaphysical (there is that word again) level. Questions are asked about the objective world and about how subjective our experience is in interaction with said world. These are very difficult questions, of which we have had to accept that there are no conclusive answers to be found. Consequently, this leads to a lack of definitive guidelines on how one should act. For if there is no clear worldview, how can we deduce our own place in the world and see where we ought to go from there?
This is where religion can come in handy, not meaning to denigrate its importance. It provides a world view and shows how this leads to the best way of acting to fulfill the optimal position of our person in this world. Christianity offers God as an explanation and the Bible as a guideline. Similarly in Buddhistic traditions, the nature of man as a reincarnate being leads to the path that brings us to Nirvana and away from suffering. Religious traditions show us where we are and where we should go.
Now, God is dead.³ When Nietszche first exclaimed this in the end of the 19th century, there was already truth to be found in that sentence. Not in the way that some omnipotent holy being passed away, but understood as pointing out the fading belief in such absolute truths as are portrayed by world religions. In the 21st century, it has become very difficult to argue with this. I would never argue that God can not exist for you, but the fact is that religions — and with that, God — have decreased in importance and the certainty of belief has faded to the extent of open discussion where there was first no doubt possible.
Where does that leave us in the modern technological world? According to Martin Heidegger, it leaves us searching for a God.
“The Need For a God”. That is a header taken straight from Hubert Dreyfus’s reading of Heidegger’s work⁴. In the first place, it seems weird to discuss the desire for a deity when reviewing one of the most important papers in the field of Philosophy of Technology. However, it is necessary to remind ourselves of Heidegger’s position concerning technology. The important aspect of technology is that we live in a world in which our “shared sense of reality”, to use Dreyfus’s exact wording once more, is what it is due to our cohabitation with technology as it exists. When we talk about technology, we talk about the world in which we live and — possibly even more importantly — the way in which we perceive said world.
In one way, this sounds dystopian, as if the technology will soon decide to end this pathetic human existence once and for all. But when we manage to accept that we live in a technological world it becomes clear that it is not a rule-or-be-ruled scenario. The answer, as very often seems to be the case, is to lean into this dependence.
The world must be understood as being filled with technologies — and Heidegger clearly states that he does not mean to revert back to an era without modern technologies — and accepted as such. Only then can we start to find out how to live with technology. For this, we must create shared values to start understanding the world through technology rather than understanding the world only as far as our technologies go. Simply put, we must put on the glasses rather than studying them.
So what is it we need? We need a shared understanding of the technological world and our relations to it. As Dreyfus puts it, we need a cultural paradigm created out of what we believe in, a paradigm that is there to stay.
We need a God.
Before we get further into what this means for us with regards to technology as a whole and Artficial Intelligence specifically, let’s take a step back. Let’s question Dreyfus and ourselves — myself — a bit here. What do we need a God for? If we know that we are inherently uncertain about what is good in a world of suffering, that there will not be a definite final answer, why should we try to create something that we can believe in?
The answer to this, is that the alternative is a pretty terrible option.
What can be seen in our world of today is that the void of not-knowing-what-is-best will be filled regardless. We are wired to strive for “better”, and for this we need a definition of what is “good”. As an additional hurdle, we have a tendency to not spend too much time thinking about this definition of good.
Consider cases in which police officers are rewarded for the arrests they make. Arrest numbers go up, which could be seen as a measure for the working of a police system. However, this does not mean that the arrests all correspond to crimes for which people should be arrested. Bringing up the arrest numbers is an easily measurable way of asserting the impact of police, but it does not necessarily measure the more complicated (and fitting) objective of bringing down crime.
This is just one example of a tendency for easy, measurable data that exists in our society. Here I am not defending that it is always a bad thing to follow what is measurable. It makes sense to do so, ethical philosophers can not run a world alone. But we should also not completely avoid the uncertainty that nuance brings with it. This is the easy way out, and what I am arguing is that this makes us lose track of what it really is that we ought to do. To make this clear(er) and (more) concrete, consider the coming to be of modern technology.
Throughout the Industrial Revolution, technologies were developed with the goal of lightening the physical workload of factory workers and enabling factories to produce more in less time. For this, efficiency⁵ was a key measure as more efficient machines led to the abolishing of more incredibly tough work conditions and to more production. Therefore, efficiency was a great means to help achieve the end goal of industrialisation.
However, the same thing happened to efficiency in technology as to arrests in police work. Over time, the means became the end. Efficiency became the way to do things. Not because this helped us in achieving a wider goal of technology, but because we started chasing efficiency itself. Still nowadays, efficiency has kept its place in the void of not knowing and it is a key characteristic goal of countless technologies.
The problem with this is that efficiency does not have a face of its own. The quest for efficiency has led to technologies that were very good in the light of efficiency, but very bad in aspects that all human ethical and lawful systems have agreed upon. Consider weapons of mass murder for example, terribly efficient and terribly wrong. Again, I am not using this example to state that efficiency is always a bad thing. However, an unbridled chase with efficiency as a single main goal is bound to produce results that we disagree with.
This is where we can find results from the absence of a God. We lack some shared sense of what ought to be our place in the technological world. As Heidegger stated, our relation to technology is more than merely instrumental. Technology is more than just used by us, it has its own character and place in our shared sense of reality. Forgetting this in our design and usage of technology, leads to a gap of not knowing how to act. Efficiency fills this gap that is left by the lack of a God in our modern world, and it is a dangerous filler.
There are more examples to be given here. Consider arguments about our health care system’s focus on merely extending lives rather than optimising the quality of lives. Think of the Toeslagenaffaire, by now a well-known example of what happens when efficiency is prioritised over ethical considerations of equality. Efficiency is a well established pillar of our modern society, and I agree with Heidegger that this is a wrong turn that we have taken on our way towards a modern society.
Alright. The lack of a God, of a shared direction, has lead us astray. It was necessary to understand where we have gone wrong to continue toward where we should go, and now the time is there to move on to the next part. Let us get into it and think about what we can focus on to improve our own relations to the technological nature of the world, away from revering efficiency and towards a more meaningful alternative.
As we discussed before, the creation of a shared cultural paradigm consists of both forming a shared sense of reality as well as creating guidelines for how we should act. Thus, we must talk about the world that we share with technology before we can discuss what it is that we should value over efficiency and how we ought to live with technology. We will first create a map to place ourselves on by talking about the nature of technology and of Artificial Intelligence specially. Afterwards, we will construct a possible direction on that map by proposing to value aesthetics — or beauty — over efficiency.
In the creation of the map lies firstly the recognition of technology as we have spoken about. A well animated summary of this can be found in this video starring Peter-Paul Verbeek, Professor of Philosophy of Technology at the University of Enschede and upcoming Rector Magnificus of the UVA. Instead of merely using technology, we must accept that we are living with technology and that there is more to this than the pure instrumental value that we perceive on the surface. Let’s get a bit deeper into this for Artificial Intelligence specifically.
In the first blog posts of this series, we talked about Artificial Intelligence and interpretation methods for those models that were considered black boxes. These methods were developed as a reaction to the impossibility of understanding the decisions made by certain deep learning models. In my opinion, this is a worthwhile field of investigation using a method which is inherently lacking a wide enough scope of investigation.
Considering the interpretability of models immediately presupposes a certain understanding of the word interpretable. It is taken to mean that we can follow the inner logic of an AI as we would follow the inner logic of a (very rational) human. This presupposition, that we can understand an artificial entity in the same way as a human, shows a flawed perception of our own relation to the entity. We see the AI as only an extension of the human behind it and we fail to understand that we are looking at a complete character which should be understood on a completely different level. To be able to move forward in understanding, we must recognise the versatile nature of Artificial Intelligence⁶.
When we start to truly consider what it is that we want from our AI’s, and where we are in relation to them, it becomes necessary to realise that Artificial Intelligence is not just intelligent. In the creation of a model, the focus has always been on how intelligent we could make said model. Such goals include finding the perfect path through highway traffic — listen carefully road users, the answer is just to stay in your damn lane — or guessing the BMI of a person by just a picture of their face. This requires intelligence, and optimising that is the goal.
However, it is inherent of a creation process that more than just intelligence ends up within the boundaries of the entity that is created from a coding basement. Like Frankenstein, a whole creature emerges upon activation. Biases, flaws, creativity, all of these characteristics can be found just as easily in artificial creatures as intelligence. We may not have focused on them in the past, but that does not mean that they were never created. We have surrounded ourselves by technology, smart technology. Now the time has come to start thinking not just of the ways in which they use their intelligence, but rather of how they act according to all of their traits. Instead of just AI, consider artificial creativity, artificial belief, artificial X.
We are starting to develop a clearer draft of the map on which we are navigating. This makes us able to look at the direction in which we are to set sail. Considering only efficiency and steering towards that has not provided us with ample guidelines for living with technology. Focusing on such simple goals to the extreme leads to dystopian scenarios. Take Brave New World by Aldous Huxley as an example, a novel in which society becomes aimed toward the abolishing of anger or sadness for its people to work efficiently and smoothly in pertaining the world’s order. This is not the aim that we are looking for, but what is?
A very important realisation here is that there is no right answer. Efficiency is not the answer, but there is no one way of valuing technology that is the best by definition. By mere deductive reasoning and / or moral debate, we are not going to be able to come to a correct answer to the question “How are we supposed to act?”. What we can do, is take into account what we know about ourselves and about the world surrounding us to propose one such way of acting. We look at existing literature, we consider what we find important as characteristics of our surrounding world, and we stay open to different perspectives when creating our own.
One such perspective is that of lecturer Yoni van den Eede⁷ from the Free University of Brussels. He argues that we should start to form what he calls an art of living with technology. One part of achieving this, is learning to value aesthetics. This sounds like an aim that is in no way more grounded than an aim for efficiency as we experience nowadays. It is certainly less scientific sounding. But this might be just what we need. As Van den Eede puts it:
“If all standard moral reasoning and ethical debate fails, or won’t suffice, an aesthetic-artful perspective — in the sense proposed here — might open our eyes to matters of ecological balance and proportion, and offer in itself a detour into fresh assessments …”
This is exactly the kind of approach that we need, in my opinion. By staying open to different assessments and acting on what we consider to be valuable in our perception of the world, we create a perspective to move forward with. Instead of merely stating the problems that come with valuing efficiency above all else — which Van den Eede does well in his book — we are given an alternative. An alternative that we are open to questioning for ourselves, but which gives a starting point for our direction in navigating the technological world in a valuable manner.
As an example for considering the aesthetic when perceiving the world through technology, consider a Spotify algorithm. Its purpose is to have us listen to more music. Is it really such an addition to our life for these algorithms to be efficient, in that they pick the music that is exactly the songs which we do not mind listening to? It might be more worthwhile to allow for machine learning to learn in a creative way, to create patterns in our music library that provide us with inspiring and beautiful changes.
This might be dreadfully annoying when it is actually played out. Maybe the fact that it is music you don’t mind listening to is what you value most in Spotify. That is also completely fine, and it is a great conclusion to come to even when it is not the same as the one I come to. However, it is dangerous to simply avoid the argument by choosing the existing efficient way of technology over a different — more aesthetic — perspective if this is done merely out of habit.
For this alone, it is already worthwhile to consider the aesthetic perspective of technology and AI and to open ourselves up to the fact that technology is never neutral. Artificial entities have characters and it is up to us to recognise this and to decide in which way we want to interact with it. It is up to us to to create a world in which we are more at home again. A more beautiful, more aesthetic world.
The danger in technology and Artificial Intelligence does not lie in Hollywood fueled dystopian future perspectives. According to Heidegger, what we should really be afraid of is losing our grasp of reality through technology. We are godless in a world that is filled with more and more complicated artificial systems and this leaves us unable to understand our own relation to technology, let alone find a meaningful way of valuing the same technology. The absence of such a God, such a shared cultural paradigm, has led us to revering characteristics devoid of meaning like efficiency. We seek for neutrality, and we find it easily measurable, but with no foundation or inherent value.
Therefore we need a God. We need to accompany our shared sense of reality with a shared sense of understanding and a shared sense of how we ought to act. To create such a paradigm is to draw a map to locate ourselves on and to decide on a direction even if this can change later. We must start by accepting our position in the technological world and by seeing the true nature of — for example — the Artificial Intelligence that surrounds us as being more than just intelligent. Afterwards, we must take a step in a direction that feels right. In this post I propose to follow Van den Eede in his pursuit of aesthetic technology, as an alternative for the established chase of efficiency. If deduction cannot decide for us, I am willing to put our fate in the hands of beauty instead.
What this exactly means, is what we are figuring out right here. It is why I am writing this blog and why I am so grateful to still have your attention after these 3.000 words of rambling. We — as participants of this experiment — are moving in the direction of aesthetic Artificial Intelligence, and that direction might well change, as we have established. As it is my belief that the value of Philosophy lies in taking a stance and navigating from this, I am willing to defend the stance of aesthetic AI. This stance might manifest in the form of more blogs (lucky you), a book, or another kind of artistic expression to see artificial entities for what they are — rather than how we have perceived them so far.
You are once again more than cordially invited to join me for the rest of the ride. To discuss with me, recommend me, or simply tell me you disagree completely. All I have to offer is my own position, some jokes in my writing and — who knows — there might be free pizza along the way.
Foot notes:
The term metaphysics comes from the ancient Greek “ta meta ta phusika”, which literally translates to “the thing after the physics”. Questions that concern the realms behind the physical one fit in well here, like “so do we live in the Matrix or not, when it comes down to it?”.
For those of you that have spent their time learning more applicable tasks, a manifold can be seen as a sheet that we fit to a certain collection of points the best way we can. In my opinion this is a beautiful metaphor and I am grateful to Janis for supplying it.
This is an often-quoted statement by Friedrich Nietszche, which could first be read (as “Gott ist tot.”) in Die Fröhliche Wissenschaft (1882).
Hubert Dreyfus’s reading of Heidegger’s The Question Regarding Technology can be found here: https://nissenbaum.tech.cornell.edu/papers/free_relation.pdf (link downloads PDF directly).
Of course, we need some sort of definition for efficiency if we are to use it as such an important part of this debate. Besides being the division of output (power) over input (power) as used in physics, we can be aided by Peter Drucker’s following quote: “Efficiency is doing the thing right. Effectiveness is doing the right thing.” Like this, efficiency is mostly referred to in management, a field I’d never thought I would find myself investigating.
This is a viewpoint that my supervisor Peter van der Putten and I have talked about before, which he was able to put into words very well. The following paragraphs are an attempt to paraphrase our discussion, and his thoughts, on the topic.
The perspective and ideas are taken from Van den Eede’s work The Beauty of Detours, which is a great work on a more complete and active stance in Philosophy of Technology.