Artificial Intelligence Is Like An Alien Species We Are Hosting On Earth
English translation of an italian post that was originally published on Levysoft.it
I recently read a long interview with a young italian startupper Riccardo Di Molfetta who, among other things, refers in a unique way to AI, comparing it to an alien species that we are hosting on Earth:
Artificial intelligence must be considered in all respects as an alien species we are “hosting” on Earth.
I believe it is wrong to think of AI as a digital version of human intelligence. The latter possesses sensitivities (imagination, vision, etc.) and experiences (emotional, cultural, etc.) that cannot be reduced to a mere computational “threshold.” Even though I believe AI could, one day, accurately simulate the mechanisms of our mind — an idea that often scares people — it will never replace the depth and uniqueness of the human spirit. To give a practical example, even when it is the machine that generates a new idea, it is the artist who decides which generated idea resonates with their aesthetic and emotional sensitivity, and how to integrate it into a personal creative vision.
In reality, I had already found a similar definition elsewhere, but more oriented towards a sociological environment, in a PhD thesis in Theoretical and Applied Social Research by Edmondo Grassi, which explores the ethical issues related to artificial intelligence. Among its 261 pages, we can read:
Artificial intelligence represents the dismantling of anthropocentrism, allowing humans to revalue their own abilities, impulses, desires, but more than anything else, their nature, as the confrontation with an alien entity could allow the exploration of fields and answering of questions that were previously insurmountable. […]
Artificial intelligence learns from human experiences, from the data that contemporary individuals provide it with at every moment of their lives, from the relationships and activities of the subject interacting with others, with what is alien to them, but only in a first stage, because that alien is nothing more than a new form of their own biology and technical-technological revolution, embracing a representation that is post-phenomenological, that is, a reality in which human-nature-technology cannot be thought of as absolute, singular, and noumenal categories, but are only different perspectives of the same model that encompasses the idea that technologies are genetic codes characterizing both the worldview and its perception by the individual, but also the very idea of being human and the relationship with a new social subject, artificial intelligence. […]
Artificial intelligences could have the task of assisting humans in discovering themselves in a more analytical, introspective, and acute way, capturing details of their nature, birth, and existence that have so far been investigated but remained without answers, thanks to the alien creature born from the cultural and evolutionary progress of the human mind and its dreams. The potential contained in the calculations, logic, and machine learning of artificial intelligence is bringing significant changes in many private and public sectors of social life, reviving concerns, ideals, defeats, and utopias of an automated world shared with another sentient being as human or close to its possibilities. Technology has always been a catalyst and processor of complex social dynamics, which seem to lead the same life, but its increasingly rooted use in society has become essential for understanding human relationships, social policies, citizens’ inclinations, and cultural and economic changes. Among the major challenges raised in the field of artificial intelligence is the impact on the world of work, unemployment, the disappearance of traditional jobs, and perhaps the birth of new ones, the increase in productivity, and the creation of new consumer goods, such as in the pharmaceutical, transportation, or entertainment sectors. […]
Artificial intelligence could contribute to the study of human relationships and their nature, in the dimension in which artificial intelligence is welcomed as an alien agent from which to learn and observe its interactions. […]
PhD thesis in Theoretical and Applied Social Research by Edmondo Grassi
Obviously, defining AI as an “alien entity” aims to represent something radically different and new compared to traditional human experiences, which can open new possibilities for understanding and knowledge previously deemed impossible, pushing humans to re-evaluate their abilities and nature. According to the thesis, this technological “alien” can help humans understand themselves better in a deeper and more detailed way. Through calculations, logic, and machine learning, AI can positively influence many aspects of human life. The vision presented here is one of a world where AI becomes an integral and collaborative part of human society, ultimately becoming a component of human biology and technological revolution, almost an extension of human nature itself. A bold vision in which the distinctions between human, nature, and technology dissolve, becoming simply different perspectives of a single complex and interconnected reality.
So, I want to quote a short video I saw on X by Been Kim discussing AI alignment and interpretability using the analogy with the translation of the Korean word “jeong” (정).
Anyone in the audience who know what “jeong” is?
Right, so I can just make it up. “jeong” is something between love and like, but it’s never a romantic thing, and also it’s something that doesn’t happen suddenly. There is no “jeong” at first sight. It has to take time.
You probably know if you speak multiple languages, you probably have words like this in your languages, too, words that cannot quite be translated between cultures, between languages, without a lot of context.
So the alignment problem is something similar. We’re trying to align values between humans and machines, except this time, it’s a more difficult question, because humans and machines, we don’t even share very basic principles, like we have families, we want to survive.
Finally, I want to mention a simple title of a book that I found in a bookstore these days:
Artificial intelligence does not think (neither does the brain)
by Miguel Benasayag, Ariel Pennisi
I haven’t read it (so I can’t say if it relates to what I’m saying) but the title alone was enough to trigger a whole series of thoughts and speculations, as if needing to reconsider one’s ideas about thought and cognition, both human and artificial, and wanting to question the traditional idea of thought, suggesting that neither artificial intelligences like ChatGPT nor human brains “think” in the way we commonly mean the term.
Before language models like ChatGPT, no software had ever demonstrated such linguistic flexibility. Not even the linguistic flexibility of a child. Now, in an attempt to understand the nature of these new models, we are faced with a disturbing philosophical dilemma: either the link between language and the human mind has been broken, or a new form of non-human mind has been created.
Futuroprossimo
