Is AI a Master or Slave?

JOSCHKA FISCHER

Source Link

BERLIN – We are living through eventful – one might even say “wild” – times, with history being made at a fast and furious pace. Why is this happening now? Because three great transformational crises have befallen humanity all at once. Each of today’s geopolitical, climatic, and digital transformations would pose enough of a challenge on its own, but we are experiencing them simultaneously. An unprecedented global mega-crisis is threatening to overwhelm our political and cultural systems’ capacities for adaptation or maintaining control.

Each passing year of record-breaking temperatures, wildfires, droughts, and extreme weather events underscores the scale of the climate crisis. Though its long-term global implications are extraordinarily complex, the basic nature of the problem is well understood. The solutions are known, but the politics for achieving them are maddeningly difficult. By contrast, the consequences of the digital transformation remain more uncertain. Just in the past year, humanity has opened a new technological door with breakthroughs in generative artificial intelligence (AI), and no one knows for sure what lies on the other side.

One big difference between these two developments is that the effects of AI could still be stopped, even reversed, at least in principle. However, one doubts that they actually will be. Both historical experience and the underlying logic of research and technological development suggest that the AI revolution will continue to gain momentum.

Toward the end of 2022, the California-based start-up OpenAI (with funding from large investments by the tech giant Microsoft) released its large-language-model ChatGPT, thereby triggering a new technological gold rush. Though AI is not new, many have recognized the current moment as the start of a new era. Digitization has reached a qualitatively unprecedented level, the thinking goes, and now will fundamentally change our whole way of life – from how we produce and consume to how we learn and interact with each other.

Looking ahead, we should consider the possibility that the relationship between humans and machines will be turned on its head. With its superior computational power and speed, access to a growing overabundance of data, and quickly improving perceptual capacities (thanks to ever-more comprehensive and sophisticated sensors), the new machines first will become indispensable for humankind, and then will become far superior to it.

What we are dealing with, then, is a potential swap between subject and object – between humans and their tools. Since machines with self-learning capabilities and superior knowledge clearly will have the ability to relegate humanity to second place, the real question is whether they will indeed do so.

To appreciate this dynamic, we should return to the German philosopher Georg Wilhelm Friedrich Hegel’s famous analysis of the “Master-Slave Dialectic,” which he articulated at the height of the first Industrial Revolution, in his seminal 1807 book, The Phenomenology of Spirit. The archetypal Master, Hegel explains, commands and enjoys the fruits of others’ labor, whereas the Slave endures the hardships of toil and subjugation to the Master’s will. In the process, however, the Slave acquires the skills for shaping the world, and one day throws off the yoke.

How will this dialectic play out in the relationship between humans and machines? This question is completely unprecedented in human history, and the answer may well determine if that history continues in the future. But complicating matters further are the other two great transformations, because these may mean that we have little choice but to embrace AI fully.

After all, is human survival on a substantially warmer planet even conceivable without establishing an AI-based technological civilization? And what are we to do if our geopolitical rivals are pursuing AI dominance, other than to pursue AI dominance ourselves? Moreover, what will become of the fundamental political makeup of human societies under these radically different and novel conditions? What will become of states and governments and the seemingly never-ending jostling between them, especially if AI becomes the preferred means of waging – or even commanding – modern warfare?

Now that we are stepping over the AI threshold – seemingly with little choice in the matter – we must take seriously the possibility that our shift from a human-centric civilization to one dominated by machines will culminate in the removal of the human element altogether. Even if the human species is not eliminated, it might be completely marginalized. Strange as it seems to us now, that is the eminently likely climax of the dialectic of enlightenment.