What this covers

Dan Shipper traces a long intellectual arc from ancient Greece to modern machine learning, arguing that Western thought has wrongly privileged explicit, rule-based knowledge over the tacit, experiential knowing that intuition provides. The conversation begins with Socrates, who demanded that knowledge be reducible to clear definition—a demand that set the trajectory for rationalism as a philosophical project and later infected artificial intelligence itself. Symbolic AI, the early approach to machine thinking that tried to codify the world as explicit if-then rules, inherited this Socratic impulse directly. Shipper shows why it failed: any attempt to define something practical—say, what makes an email important—runs into an endless cascade of exceptions, each rule generating the need for further rules, until the project collapses under its own weight.

Neural networks succeed where symbolic systems foundered because they learn the way humans actually do: through thousands of examples and pattern-matching, not through explicit definition. They operate like intuition, learning inexplicit rules that can be observed in their behavior but never itemized as a complete checklist. Shipper argues this technological turn vindicates what Protagoras was defending in his ancient debate with Socrates—that stories, myths, and hands-on experience are genuine ways of knowing, just as valid as what can be explicitly stated. The intervening centuries, shaped by Enlightenment thinkers who married rationalism to mathematics, rendered intuitive knowledge invisible. Shipper contends that rationalism has worked brilliantly in physics but poorly in fields like psychology, where the replication crisis suggests that reducing human phenomena to universal laws may be fundamentally misguided. His thesis is that we need both modes—intuitive framing and logical manipulation within it—and that modern neural networks finally make this balance visible again.

Sharpest takeaway

Shipper argues that rationalism—the demand that knowledge be made explicit, rule-based, and definable—is only half the story, and that neural networks, by working like human intuition, finally vindicate the tacit, experiential ways of knowing that Socrates dismissed.

  • Symbolic AI failed because reducing the world to explicit if-then rules proved too brittle and exception-ridden
  • Neural networks succeed precisely because they learn inexplicit patterns from examples, mirroring human intuition
  • The Socrates-Protagoras debate is the origin point where Western thought equated knowing with being able to define

The claims · ranked20 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.86

Any attempt to define something like an important email through explicit if-then rules inevitably runs into an endless cascade of exceptions—each rule requiring further rules to handle abuse and edge cases—so that defining one concept eventually requires defining nearly everything about the world, making the project of total explicit definition too brittle and computationally infeasible to work.

causalhigh valueestablishednovelty 3/4durability 4/4· Dan Shipper

“anytime you try to create rules to define these things, you always run up against exceptions”

0.80

Rationalism is the idea that to truly know something you must be able to describe it explicitly—reducing it to theories and rules for how the world works—and that this explicit knowledge is what counts as true knowledge, distinct from everything else.

definitionhigh valueestablishednovelty 2/4durability 4/4· Dan Shipper

“Rationalism is really the idea that if we can be explicit about what we know, if we can really reduce what we know down into a set of theories, a set of rules for how the world works, that is true knowledge about the world”

0.78

Our best metaphors for our own minds tend to be whatever tools we are using at the time—Freud modeled the mind on the steam engine, and in the 20th century the dominant metaphor became the computer, leading us to aspire to be logical, rational machines and to make the importance of intuition invisible.

causalhigh valuecontestednovelty 3/4durability 4/4· Dan Shipper

“our best metaphor for our minds are the tools that we use. So a really good example is Freud has one of the most impactful models of the mind... he used the steam engine as a metaphor”

0.78

Stories, myths, metaphor, and personal hands-on experience give us a genuine way of knowing things we may not be able to explicitly state, and we know those things just as much as things we can explicitly say; this is what Protagoras was defending and what was lost in the birth of rationalism when Socrates insisted that what cannot be defined is not known.

normativehigh valuecontestednovelty 3/4durability 4/4· Dan Shipper

“stories and and personal hands-on experience give us a way of knowing about things that we may not be able to explicitly talk about, but we still know just as much as we know things that we can explicitly say”

0.78

Rational thought emerges out of intuition: we first have a squishy, inexplicit intuitive understanding of the world that sets the frame, and then rational, methodical, logical thinking goes in and manipulates things within that frame—so we genuinely need both modes, not just the rational one.

causalhigh valuecontestednovelty 3/4durability 4/4· Dan Shipper

“you can think of rationality as emerging out of intuition”

0.78

Herbert Simon and Allen Newell, two founding fathers of AI, built the 'general problem solver,' which because computers were so expensive was first codified on paper and executed by hand—one of them reportedly having his family help simulate how a computer would solve complex problems.

factualhigh valueestablishednovelty 3/4durability 3/4· Dan Shipper

“the two founding fathers of AI, Herbert Simon and Allen Newell, built this machine that they called the general problem solver”

0.76

Symbolic AI worked well on simple toy problems but failed as problems grew complex because the search space of possible solutions became enormous; representing real-world problems as logic problems caused the systems to break down once they moved beyond toy cases.

causalhigh valueestablishednovelty 2/4durability 4/4· Dan Shipper

“as problems got more and more complex, the search space of possible solutions got really, really, really, really big”

0.73

While rationalism has been wildly successful in physics, it has been very hard to make comparable progress in fields like psychology, economics, and neuroscience, which are structured in imitation of physics by trying to reduce complex phenomena to definitions and universal laws—an approach evidenced by psychology's ongoing replication crisis after a century of research.

causalhigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“in areas of the world like psychology or economics or neuroscience, it has been really hard to make progress in the same way that physics has made progress”

0.73

Early AI theorists adopted Symbolic AI—the idea that thinking could be embodied in logical symbols and explicit rules for transforming them—which forms a clear through line from Socrates's project of reducing knowledge to explicit definitions through the Enlightenment to AI's founding at Dartmouth in the 1950s.

factualhigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“there's a really clear through line from Socrates's original project to the Enlightenment to the original approach that AI theorists took called Symbolic AI”

0.73

The birth of rationalism can be traced to Plato's dialogue 'Protagoras,' where Socrates demands an explicit, non-contradictory definition of excellence while Protagoras argues via myth and story that everyone has the capacity for excellence; by exposing that Protagoras cannot define it, Plato implies he does not know it, setting Western thought on the path of equating knowing with defining.

factualhigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“you can trace the birth of rationalism to this dialogue, "Protagoras." And in the dialogue, it's a debate between Socrates on the one hand and Protagoras”

0.73

Language models learn thousands of partially-fitting, inexplicit rules to predict the next word; these rules can be observed in the network's behavior but do not exist as a definitive listable set anywhere inside the network—just as you could not find an explicit list of rules in a human brain under a microscope.

factualhigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“all of those rules are inexplicit. They're kind of like, you can observe them in the overall behavior of the network, but they don't exist anywhere in the network”

0.71

A neural network, inspired by the brain, consists of layers of connected artificial neurons that learn to recognize patterns by being given many examples and being corrected when wrong; over many iterations it learns to perform complex thinking-style tasks without any explicit set of definitions or rules, succeeding where symbolic AI failed.

definitionhigh valueestablishednovelty 2/4durability 3/4· Dan Shipper

“what you can do with a neural network is you can get it to recognize patterns by giving it lots of examples”

0.69

Enlightenment thinkers like Descartes, Newton, and Galileo took the Socratic idea that knowing requires explicit description and turned it into a method of using mathematics to explain and predict the world, reinforcing that to truly know something you must describe it explicitly and ideally mathematically.

factualhigh valueestablishednovelty 1/4durability 3/4· Dan Shipper

“Thinkers on the philosophy side, like Descartes, and on the science side, like Newton and Galileo, took this idea and used it as a new method to understand and explain the world.”

0.68

Socrates is the father of rationalism because he was one of the first to insist that knowledge requires being able to clearly and explicitly define what we know and how we know it, setting Western society on a path of identifying knowledge with the ability to define.

factualhigh valueestablishednovelty 2/4durability 3/4· Dan Shipper

“I think the father of rationalism is Socrates the philosopher. Socrates is one of the first people to really examine the question of what we know and how, what is true and what's not true.”

0.68

History shows it is better to be open to more ways of seeing and working with the world, and in this particular era it is especially important to be able to work with things that are a little bit mysterious and to be comfortable with that ambiguity.

normativehigh valuecontestednovelty 2/4durability 3/4· Dan Shipper

“History shows that it is better to be open to more ways of seeing and working with the world and that in this particular era, it's very important to be able to work with things that are a little bit mysterious and be comfortable with.”

0.66

Neural networks are the first technology ever invented that operates much like human intuition, since both are trained by thousands of hours of direct experience and example rather than explicit rules; this should make more visible to us the value and importance of intuitive thought that rationalism rendered invisible.

causalhigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“neural networks are the first technology we've ever invented that works a lot like human intuition”

0.66

We may not understand the particulars of how human minds or language models reach their conclusions, but that does not mean we don't understand them—we understand them in a different, intuitive way, such as developing through repeated use an intuition for when ChatGPT is good, bad, or hallucinating, just as we sense when a friend is sad or being untruthful.

normativehigh valuecontestednovelty 3/4durability 3/4· Dan Shipper

“if you use ChatGPT all the time, you develop an intuition for what it's good at and what it's not good at and when it might be hallucinating”

0.49

Symbolic AI succeeded only in limited domains, such as the expert systems of the 70s and 80s that tried to reduce medical diagnosis to a set of rules; even there, reducing diagnoses like measles or anxiety to explicit rules proved complicated and impossible to get right 100 percent of the time.

factualestablishednovelty 2/4durability 3/4· Dan Shipper

“there were these things called expert systems, for example, in the 70s and 80s that tried to, for example, reduce medical diagnosis down to a set of rules”

0.34

A lot of therapy is about rationalizing or rationalizing through what you think and what you feel, reflecting how deeply rationalism is built into everyday practices.

factualcontestednovelty 1/4durability 2/4· Dan Shipper

“Even the way that we do therapy, a lot of therapy is about rationalizing or rationalizing through what you think and what you feel.”

0.30

There is no universally agreed-upon definition of AI, mirroring the broader struggle to define what it is to know something or to define concepts like anxiety in psychology; AI's project is to build a computer that can think and learn the way humans do.

factualestablishednovelty 1/4durability 2/4· Dan Shipper

“there's no universal agreed upon definition for this in the same way that we've struggled to come up with a universal definition for what it is to know something”