Listen to our fears

A significant part of the left is arguing that AI has displaced the class struggle into the digital realm. I am surprised by this stance, because the left has been characterized by a very marked technological mentality. How many times have we heard that philosophers have only interpreted the world, when what is needed is to change it? The assumption that supports this thesis of Marx is that the world is an artifact that can be reconstructed in harmony with our ideals. In this conviction lies the essence of technology.Today we consider change a priority, not because it is in our hands to modify the direction of history, but because it has become independent of our will. Those who do not adapt to the new, futurists tell us, are condemned to disappear.When Marx wanted to change the world, he behaved like a representative of truth. Today, truth is change. That is why there is no way to coherently criticize AI from the left without radically changing the image of the world as a manipulable object; that is, without recovering a relationship with the world that is not exclusively utilitarian and that allows us to stop and listen to our own fears. This simple gesture of stopping to listen to the surrounding world is, in itself, an affirmation of freedom that allows us to anchor our urgencies in the riverbed of Heraclitus' river. We do not know if we are boarding the Titanic, but we have assumed that the only thing we cannot do is not board, because there is no longer solid ground. We do not know, either, if the icebergs are illusions of our own fear. We are no longer able to differentiate what is real from what is possible.

I will give you two examples.In 2003, Nick Bostrom, a philosopher at the University of Oxford, proposed the unsettling thought experiment of the paperclip maximizer. Its goal is to help us think about the potential risks of a highly evolved artificial intelligence programmed to unconditionally pursue a seemingly harmless, but unlimited, goal.Imagine we ask it to make paper clips in such a way that it continuously maximizes its production. Remember we are dealing with a superintelligent AI that knows how to autonomously find creative strategies to achieve its goal, starting with the constant improvement of its capabilities. Consequently, it will prevent interference from anything that hinders clip production. There will be no way to turn it off. The more clips it makes, the more resources it will need, and consequently, the more it will reinforce anything that can help it increase production. This AI would not be an enemy of humans, but neither would it be their friend. It would limit itself to considering them potentially usable material.More recently, economists Brett Hemenway Falk and Gerry Tsoukalas have concluded that AI could be leading us to unlimited productivity with zero demand. It would produce in increasing quantities what no one would be able to buy. Every time a company replaces a worker with AI, it seems to act rationally... But workers are consumers and, if they are laid off, they will reduce their consumption. Companies will then cut costs by further automating production, which will favor the fall in demand and they will venture into a cycle that we do not know if it would have a natural exit.According to these researchers, neither universal basic income, nor income taxes, nor job retraining programs, etc. can prevent the growth of this productive “rationality”. The only thing that, a priori, might seem effective would be the imposition of a tax on automation, so that a company has to pay, in the form of a tax, what it saves by laying off a worker. But no one is discussing this alternative. Are companies then condemned to follow a productive rationality that has no possibility of stopping it?Do these two examples show the real, the possible, or the chimeric? It is becoming increasingly difficult to align our imagination (what is possible) with our production (what is real). Despite everything, some questions seem to arise:Can AI be criticized with a technological mindset that sees the world as an artifact within our reach?If the objectives we set for AI are not well defined, will we not take considerable risks?Is aligning AI systems with human values a technical or moral challenge?Are we heading towards the obsolescence of man?How can a world that moves at full speed and without brakes be changed?