Article Summary
This article discusses Alan Turing’s famous assumptions about artificial intelligence, which may have led research down the wrong path over 75 years. Computer scientist Peter J. Denning presents a critical view that emphasizes the limitations of current artificial intelligence in grasping tacit knowledge, which characterizes human minds. The article also points to the challenges that arise when trying to represent intelligence, cultural interaction, and context in machines, and it conveys warnings about the potential risks of advanced artificial intelligence systems that do not deeply understand the human context.
Turing and Artificial Intelligence: Old Assumptions and Complex Paths 🌍✨
Alan Turing, the British scientist considered one of the pioneers of modern computing, has long formed the cornerstone of research in the field of Artificial Intelligence. In 1950, Turing put forward two main assumptions: the first assumes that intelligence can exist independently of the physical body, and that it is possible to simulate it programmatically. The second is the idea of the Turing test, in which a machine is considered intelligent if it succeeds in imitating a human in conversation.
Over the decades that followed, thousands of studies and technologies were based on these assumptions. But were these foundational principles completely correct? It is now being considered that some of Turing’s assumptions may have led us down the wrong path in developing artificial intelligence.
Tacit Knowledge: The Invisible Barrier to Artificial Intelligence 🧭
In his new book “Turing’s Mistake: Escaping the Yoke of Unintelligent Machines,” Peter J. Denning focuses on the problem of tacit knowledge, which is the large share of knowledge that humans possess but cannot express directly or in clear digital form.
This knowledge includes:
- Common sense that a person uses in daily interaction.
- Social and environmental interactions that emerge in real-life situations.
- Feelings and sensory perceptions, which deeply affect decision-making.
- Specialized practical skills that require a body and long experience (for example, a skilled violinist).
- Social and cultural knowledge that characterizes societies and frames events.
This open-ended body of knowledge is extremely difficult to insert into machine-learning algorithms or complex databases, raising questions about the possibility of a complete copy of human intelligence in machines.
Previous Experiments: The Cyc Project and Common-Sense Gathering
In the 1980s, Douglas Lenat tried to develop the “Cyc” project as an experiment to collect as many facts and truths as possible that represent common sense within a huge database. After years of work, the project’s database reached nearly 25 million facts.
However, as Denning points out, even this enormous amount was not enough to make systems “intelligent” in the human sense. The real expertise behind complex decision-making includes things that cannot be expressed by sets of facts alone.
Practical Skills and Emotional Perception: The Embodiment Barrier 👩🎤🎭
Skills such as playing music, performance arts, or even the ability to use tools skillfully, remain far from complete digital representation. A violinist may produce stunning music but cannot fully explain how he creates the beauty of sound, and this is what Denning calls “embodied knowledge.”
Even if robots can imitate human behavior, they lack the ability to feel and experience things physically and emotionally, which is a fundamental difference between human and machine intelligence.
Context and Culture: The Missing Key to Human Understanding 📸🧭
The challenges do not stop at tacit knowledge alone, but extend to the context and culture that shape human intelligence. Dealing with sarcasm, jokes, emotional expressions, and sensitivity to social circumstances are all things the machine finds difficult to capture accurately.
Denning stresses the importance of context and how words or actions derive their meaning from backgrounds and from conversations that are linked in a complex way. This “cultural fabric” covers values, beliefs, and customs that differ between peoples and even between small communities.
Accordingly, attempts to build large language models (Large Language Models) as in ChatGPT and others do not enable these models to go beyond their limits as mere processors of words, without truly possessing a concept or a deep understanding of what they say.
Global Risks: Do We Not Know What Machines Are Making?
Denning warns that machine intelligence will evolve in different directions from human intelligence, creating a gap that will be difficult to bridge between humans and machines. Major concerns arise about AI Safety, because machines may continue performing intelligent tasks while not understanding human motives or interacting with human intentions in the required manner.
The result may be the emergence of systems that are “intelligent” in a narrow domain but have the ability to cause major problems for humans in terms of technical or ethical dangers, and thus the challenges become greater than merely the potential dominance of superintelligent machines.
Coexisting with Artificial Intelligence: A Call for Rethinking 🤖🌍
In light of this new vision, Denning calls for the need to reassess the way we think about artificial intelligence:
- Reject accepting the premise that machines must be an exact replica of human intelligence.
- Reaffirm the human differences that include physical experience, emotions, culture, and context.
- Build a future that consciously deals with the emergence of smart technologies that we may not fully understand, while emphasizing the priority of the human element.
Conclusion
Although Alan Turing opened wide doors to the concept of artificial intelligence, contemporary developments show that some of his assumptions now face fundamental challenges. Tacit knowledge, cultural context, and deep human interaction all constitute important obstacles to achieving artificial intelligence that equals human intelligence. While technology advances at a rapid pace, the most important question remains how to coexist safely and in balance with these devices that have become part of our daily world.
With this perspective, critical thinking offers an opportunity to redirect the course of artificial intelligence research toward better integration with our true human understanding, not merely surface-level imitation.
This article reflects a modern critical view that addresses the challenges of artificial intelligence inspired by revisiting Alan Turing’s ideas after more than seven decades.
📚 Related articles
Discover more from Mohdbali
Subscribe to get the latest posts sent to your email.
