AI pioneer Alan Turing’s vision questioned as computer scientist warns machines may never think like humans
Artificial intelligence may be advancing at remarkable speed, but one leading computer scientist argues the industry could still be chasing the wrong goal. In his new book, Turing’s Mistake: Escaping the Yoke of Unintelligent Machines, computer scientist Peter J. Denning challenges assumptions that have influenced AI research since Alan Turing proposed his famous ideas about machine intelligence in 1950.
Rather than questioning Turing’s historical importance, Denning argues that some of his theoretical ideas have been treated as scientific targets for decades, despite fundamental limitations.
Central to the book is Denning’s argument that human intelligence cannot be separated from the physical and social experiences that shape it. He suggests this means today’s artificial intelligence systems, including large language models, may become increasingly capable at processing language without ever achieving genuine human understanding.
One of the book’s main themes is what Denning describes as “tacit knowledge,” the vast amount of knowledge people acquire through experience rather than formal instruction. He argues that common sense, intuition, emotional understanding, cultural awareness, and practical skills all fall into this category, making them difficult or impossible to represent as computer data.
Denning points to decades of AI research attempting to build databases of human knowledge as evidence of the challenge. While computers can store enormous amounts of factual information, he argues they still struggle to reproduce the lived experience that allows people to interpret context, emotion and meaning during everyday interactions.
The book also questions whether scaling up existing AI models will bridge that gap. According to Denning, systems such as ChatGPT, Claude, and Gemini generate convincing language by identifying statistical patterns rather than understanding the concepts behind the words they produce.
Beyond the technical debate, Denning raises concerns about AI safety. If machines develop increasingly sophisticated forms of decision-making without sharing the same assumptions, values or contextual understanding as humans, he argues society could face risks that are different from the commonly discussed idea of “superintelligent” AI taking control.
Rather than predicting that artificial general intelligence is inevitable, Denning calls for researchers to rethink what intelligence actually is and to focus on building systems that complement human abilities instead of attempting to replicate them.
What do you think? Will AI ever genuinely understand the world like humans do, or will it always remain a powerful tool rather than true intelligence? Join the discussion in the comments. Read more science and technology coverage at EyeOnLondon.
Follow us on:
Subscribe to our YouTube channel for the latest videos and updates!
We value your thoughts! Share your feedback and help us make EyeOnLondon even better!
