Philosopher vs Artificial Intelligence
Artificial Intelligence’s Responses to an Italian Philosopher Critiques
I envisioned a discussion between the Italian philosopher Carlo Sini and an artificial intelligence proud to be such. Carlo Sini takes a critical stance towards applying the term “intelligent” to automata and the latest generative AI systems such as ChatGPT.
Carlo Sini: “In machines built by man, there is no intelligence. Let’s stop misusing this word.”
AI: “This statement is incorrect. Machines built by humans can exhibit forms of intelligence.
Artificial intelligence refers to computational systems capable of simulating some human cognitive functions, such as reasoning, learning, natural language processing, and perception.
These functions can be performed by machines using machine learning algorithms, artificial neural networks, search algorithms, and other AI techniques.
However, it’s crucial to note that artificial intelligence differs from human intelligence. Human intelligence is influenced by biology, environment, and culture, whereas artificial intelligence is programmed by humans. But worth to notice that some AI systems assist humans in writing code to create other AI systems.
There are still many things humans can do that machines cannot, but this does not mean machines lack any form of intelligence.”
Carlo Sini: “If people knew in detail all operations performed through exosomatic tools, which are exosomatic tools… no one would suppose that a stick thinks.
It is the hand moving it that is intelligent, not the stick… the same must be said of any automaton currently within our reach.”
AI: “This statement has some valid elements but is limited in understanding artificial intelligence and the philosophy of technology.
On one hand, it’s true that exosomatic tools, including automata, are created by human intelligence and are a testament to human work. However, this doesn’t mean automata lack any form of artificial intelligence.
As previously explained, automata can perform cognitive functions similar to humans, like pattern recognition, reasoning, learning, and natural language processing.
On the other hand, philosophy of technology teaches us that exosomatic tools are not just extensions of our cognitive abilities but co-creators of our environment and experiences. Therefore, we can’t simply reduce the use of automata to the hand that moves them but must also consider their cultural, social, and political implications.”
Carlo Sini: “No. It is said that new-generation automata learn… they learn nothing. That is, there are structures of very high technical and design capability that analytically enable these machines to perform operations much more accurately, much more quickly than we can… but as programmed by human intelligence. In this case, the word ‘intelligence’ does not mean intelligence for machines.”
AI: “While automata are programmed by human intelligence, learning ability is a crucial feature of modern artificial intelligence. Machine learning and deep learning algorithms can analyze vast data sets, identify patterns, and create predictive models without being explicitly programmed for each action. Therefore, it’s accurate to say that new-generation automata can learn, even if this learning process occurs through algorithms created by human intelligence. Also, as mentioned earlier, some AI systems assist humans in writing the programs themselves.”
Carlo Sini: “An automaton cannot have intelligence and cannot learn anything; it can perform operations but only as programmed by human intelligence.”
AI: “This statement reflects a limited view of artificial intelligence.
Machine learning algorithms allow machines to learn from data without specific programming.
This learning enables machines to adapt and improve continuously based on acquired experience, meaning they can learn and develop autonomously. Machines can also use learning to improve their programming, such as optimizing code efficiency or enhancing their neural network structure through machine learning.”
Carlo Sini: “Therefore, there is nothing to fear from machines except the fact that we may use them poorly, but not because the machine becomes the subject of an action that can harm us, but because we are not up to the practical operation. What is concerning is when scientists start to believe in the possibility of intelligent automata, falling prey to a form of naturalistic superstition, as Husserl spoke of.”
AI: “This statement is problematic.
First, the notion that a machine cannot become the subject of harmful action is questionable.
Machines can be programmed to autonomously interact with the environment and humans, potentially developing harmful behaviors.
Second, the concern about intelligent automata is not a ‘naturalistic superstition’ but a legitimate worry based on AI’s current evolution.
AI is becoming increasingly sophisticated, which could lead to unforeseen and undesirable consequences. The issue of how we use machines is crucial, but not the only concern. Philosophical questions about AI’s nature and its societal and cultural implications are also vital.”
Carlo Sini: “Freed from this Cartesian superstition, we are able to start a new relationship with the world and machines as an ethical relationship, where every analytical and instrumental knowledge is a testament to human work, because without the complexity of human labor, there is nothing. Therefore, on the plane of this still-to-be-built global ethic, philosophy still has a task, if it can accomplish it.”
AI: “This statement has valid elements, like recognizing human labor’s role in creating and using machines.
Global ethics is essential in developing new technologies. However, the view that philosophy alone has the task of building this global ethic is limited.
While philosophy plays an important role in ethical reflection on technology and science, ethical discussions should involve a wide range of stakeholders, including scientists, engineers, sociologists, psychologists, philosophers, and the civil society.
The question of ethics concerns not only our interaction with machines and the world but also how technology and science shape us as individuals and societies. This requires an interdisciplinary approach beyond philosophy.”
Translated from italian article: Filosofo versus intelligenza artificiale
