Which Historical Figure Would Have Been Most Intrigued by Today’s AI?
This content was generated with AI assistance.
Why Leibniz Would Have Been Most Intrigued by Today’s AI
My strongest candidate is Gottfried Wilhelm Leibniz (1646–1716)—more than Alan Turing, Ada Lovelace, Leonardo da Vinci, or even Benjamin Franklin.
Leibniz would not merely have regarded ChatGPT as an astonishing machine. He would probably have seen it as the first crude, unreliable manifestation of a project he had been imagining for most of his life.
Why Leibniz?
Leibniz wanted to build something resembling a universal system of knowledge and reasoning. His vision had several interconnected parts:
- a universal representation for human concepts;
- an encyclopedic store of organized knowledge;
- a method for combining concepts and deriving conclusions;
- machines that could perform portions of intellectual work.
He called the proposed representational system the characteristica universalis, roughly a universal language of thought. Alongside it was the calculus ratiocinator, a system in which reasoning could be performed through formal operations rather than endless verbal argument.
His famous underlying ambition was essentially:
Stop arguing vaguely. Represent the ideas properly, then calculate the answer.
An LLM is not his envisioned logical calculus—it is probabilistic, linguistically messy and capable of inventing facts—but the experience of using one comes remarkably close to his broader dream:
- Much of recorded human knowledge has been compressed into a machine.
- You address that machine in ordinary language.
- It translates among fields, languages and representational forms.
- It combines existing ideas into apparently novel formulations.
- It can help reason, write, explain, classify, summarize and devise plans.
- It acts as a generalized intellectual instrument rather than a single-purpose calculator.
That last point is decisive. Leibniz was a compulsive unifier. He moved among mathematics, philosophy, logic, law, language, engineering, geology, theology, diplomacy and librarianship. ChatGPT would not look to him primarily like “a chatbot.” He would recognize it as an embryonic universal intellectual interface.
His reaction would probably occur in stages
First: complete astonishment
The fact that a machine could discuss Roman law, calculate an equation, translate Chinese, explain metaphysics, draft a letter and propose a mechanical design in one continuous conversation would strike him as evidence that his life’s central intuition had been correct: knowledge can be represented and mechanically transformed.
Leibniz also worked seriously on binary arithmetic and calculating machines, making the connection to digital computation unusually direct.
Second: frantic experimentation
Within hours, he would likely be trying things such as:
- comparing translations of philosophical terminology;
- constructing formal taxonomies of knowledge;
- generating combinations of scientific hypotheses;
- converting legal arguments into structured representations;
- asking it to reconcile apparently contradictory philosophical systems;
- using it as an assistant for correspondence, research and invention;
- attempting to build a better symbolic reasoning layer around it.
Leibniz produced an enormous volume of correspondence and unfinished projects. He would understand immediately that an LLM is partly a cognitive bandwidth multiplier: it allows one energetic polymath to pursue far more threads simultaneously.
He would be insufferably productive.
Third: profound irritation
He would quickly discover that LLMs do not actually implement his calculus ratiocinator.
They can:
- sound certain without being correct;
- blur distinct concepts;
- hide their inferential steps;
- produce different answers to equivalent questions;
- substitute plausible wording for valid deduction.
Leibniz wanted reasoning that was explicit, inspectable and mechanically verifiable. Contemporary LLMs often provide the opposite: astonishingly capable reasoning wrapped in an opaque probabilistic process.
His response probably would not be rejection, however. It would be:
This is magnificent. Now we must make it rigorous.
He would immediately begin designing a hybrid of:
- LLMs for language and conceptual exploration;
- formal logic for deduction;
- databases and knowledge graphs for factual grounding;
- theorem provers for verification;
- provenance systems for tracing claims;
- agentic tools for taking actions.
In other words, he would probably propose something remarkably close to where advanced AI systems are now heading.
Why not Alan Turing?
Turing is the obvious answer, and he would certainly have been fascinated. He explicitly considered whether machines could exhibit intelligent behaviour through conversation.
But Turing’s likely reaction would be more analytical:
- Does it genuinely reason?
- What behaviour does it exhibit?
- What experiments distinguish imitation from intelligence?
- How does the mechanism work?
Turing might be the person most intellectually vindicated by ChatGPT.
Leibniz would be the person most delighted by its possibilities.
That distinction matters.
Why not Ada Lovelace?
Ada Lovelace understood that computing machinery could operate on symbols and potentially manipulate music or other non-numerical structures. She would immediately grasp the importance of generative AI.
But she was also cautious about attributing true originality to machines. LLMs would force her to revisit that position in fascinating ways.
She might be the best historical figure to debate whether AI is creating or merely elaborating relationships already implicit in its training.
Why not Leonardo da Vinci?
Leonardo da Vinci would use multimodal AI with almost reckless enthusiasm—image generation, mechanical design, anatomy, architecture and scientific visualization.
But Leonardo did not formulate as clear a philosophical programme for mechanizing knowledge and reasoning. He would be an extraordinary user of AI.
Leibniz was anticipating something much closer to its conceptual architecture.
Why not Jorge Luis Borges?
Jorge Luis Borges deserves an unusual honourable mention.
A model trained on enormous portions of written culture—capable of producing endless variations, invented citations, plausible nonexistent books and labyrinthine transformations of language—would feel as though it had escaped from one of his stories.
He might understand its cultural and metaphysical strangeness better than anyone:
- a library compressed into a statistical object;
- an author without experience;
- a text generator haunted by every text;
- a machine that remembers patterns but not events;
- an oracle that cannot reliably distinguish truth from plausibility.
Borges might be the historical figure least surprised aesthetically by an LLM.
Final ranking
- Leibniz — most excited overall
- Turing — most scientifically vindicated
- Ada Lovelace — most perceptive about symbolic creativity
- Borges — most attuned to its philosophical weirdness
- Leonardo — most enthusiastic power user
My conclusion is therefore:
Leibniz would see modern AI as the incomplete birth of the universal intellectual machinery he had spent his life trying to imagine.
And unlike some historical thinkers who might simply marvel at ChatGPT, Leibniz would almost immediately understand the next problem: language generation is not enough; it must be joined to verified knowledge and reliable reasoning.
That is still one of the central unsolved engineering problems in AI.