In 1952, Alan Turing published his final work, in which he acknowledged that intelligence exists in various forms. He explored morphogenesis, the process by which we evolve from a single cell into complex organisms. This remarkable process somehow directs our cells, which function like tiny computers running elaborate genetic programs encoded in our DNA. Turing demonstrated that specific equations, resembling chemical reactions, can naturally generate patterns.
In the United States, several notable scientists were also contemplating the concept of intelligence. Among them, the most renowned was the mathematical prodigy John von Neumann. He played a pivotal role in developing the world’s first programmable computer and authored a highly influential report that led to the establishment of the von Neumann architecture, which all subsequent computers were said to emulate.
Von Neumann examined a phase of biology that predated Turing’s work: how does life replicate itself? He entertained an idea that still resides in the realm of science fiction today. Picture a sophisticated machine capable of self-replication. Max Tegmark discusses this in his book Life 3.0, predicting that during the 21st century, we will witness the emergence of such machines—intelligent machines creating other smart machines.
However, von Neumann was unable to construct such a device. Contrary to many claims made by the technology sector, we still lack the physical technology necessary to realise this ambition. Von Neumann’s concept of a universal constructor existed only in theory, yet it represented an extraordinary feat of engineering. He devised the concept of cellular automata, a grid of cells in which each cell can assume one of 29 states. The state of each cell is influenced by the states of its neighbouring cells. His design was remarkably ingenious; the self-replicating machine would use a set of instructions to create a new copy, which would then transfer those instructions to the new machine, enabling it to replicate itself.
Years after von Neumann’s passing, researchers continued to explore this idea. It wasn’t until the 21st century that computer scientists successfully demonstrated von Neumann’s designs operational on modern computers.
As we have seen in this and the two previous blogs, the origins of AI do not align with the commonly accepted year of its beginning—1956. The late 1940s and early 1950s were already notable for individuals committed to creating machines capable of human-like thinking. Some referred to this as cybernetics, while others preferred terms such as information processing or automata theory. In 1955, a group of American scientists proposed the first conference to gather everyone interested in this burgeoning field of research, to be held at Dartmouth College. In their proposal, they designated the field as Artificial Intelligence. The conference took place in 1956 and lasted eight weeks, laying the groundwork for many key concepts that would dominate the field for years to come. We will explore several of the approaches discussed at this conference in upcoming blogs.