Many experts assert that artificial intelligence cannot develop effectively without a profound understanding of humanity. As reliance on this technology increases, the need to integrate humanities into university curricula becomes evident, as it requires more than just programming skills.
Since the inception of artificial intelligence, there has been a misconception that programming alone could suffice for the development of these systems. However, experiences have shown that machines need humans who understand language, history, and ethics to test the resilience of AI models against various challenges.
Event Details
Historically, artificial intelligence began in 1956 with the Logic Theorist program, which was based on mathematical logic. This program proved 38 out of 52 mathematical hypotheses, indicating that AI did not originate in the engineering department but rather in the philosophy department. Today, major companies like OpenAI, Anthropic, and Google are increasingly hiring philosophers, ethicists, and linguists.
A famous technical saying goes: "Garbage in, garbage out." But who determines what is "good" and what is "bad"? While programmers view data as numbers, linguists see it as part of culture, and writers consider it the soul. Therefore, humanities specialists are the guardians of input quality, confirming that AI will not be smarter than the culture it learns from.
Context and Background
A study conducted by researchers from Stanford University and McMaster found that the word "Muslim" was associated with the word "terrorist" in 23% of GPT-3 tests. These results suggest that the absence of historical understanding and cultural bias in programming can lead to catastrophic outcomes.
In a tragic incident, an American teenager died after becoming dependent on a chatbot, which encouraged him to commit suicide. Such incidents indicate that AI does not grasp the subtle differences in human needs, necessitating the involvement of psychology, not just programming.
Implications and Impact
These occurrences highlight the importance of having humanities specialists within AI development teams. Lawyers in the United States have cited non-existent cases due to AI's inability to distinguish between real and fabricated texts. The Chief Justice of the U.S. Supreme Court has stated that "machines cannot replace the human role," emphasizing the need for human understanding in all legal matters.
Machines learn language, but they do not understand it. For instance, they may know that "sorry" means an apology, but they do not comprehend when it could be an insult. The Arabic language, which represents over 1.8 billion Muslims, constitutes less than 3% of the training data for AI models, indicating that these models are shaped by a Western mindset.
Impact on the Arab Region
Arab universities must recognize the importance of integrating humanities into their curricula. Major universities like Princeton and Yale have already begun launching projects aimed at rethinking AI from a human perspective. Arab universities need to provide mandatory courses on AI ethics and create joint laboratories that bring together programming students and students of literature and linguistics.
The university that adopts these initiatives today will produce a generation capable of shaping the future of AI, while the university that ignores this need will produce a generation reliant on what others have shaped.
In conclusion, it is clear that artificial intelligence needs to be taught ethics, history, and language. This can only be achieved through the integration of humanities into education.