English | EN 
Patrick Morley took the stage with a teacher’s dry wit and a statistician’s precision, introducing himself as “a recovering educator still amazed that students survive the nonsense we put them through.” From that start, it was clear this wouldn’t be another tech evangelist talk. His subject—the probabilistic nature of modern AI— wasn’t about machines at all, but about the way we think, err, and occasionally fool ourselves
From Logic to “Definitely Maybe”
Morley structured his keynote into three acts: The Evolution of Thought, The Mathematics of Thought, and How to Regain Agency. The connecting thread? Bayes’ Rule, a formula he called “two words that define all of AI: definitely maybe.” He reminded the audience that human thinking has only been logical for about 1% of our species’ history. For the other 99%, we relied on myths, stories, and gut feeling. “Mythology,” he quipped, “is like an embedded system running in the background.” Tracing the lineage of thought from Xenophanes— who noticed that people made gods in their own image—to Plato, who gave us the questions Why, What, How, and Aristotle, who codified logic into rules, Morley showed how reasoning evolved from storytelling to structure. Aristotle’s famous syllogisms — “All men are mortal; Socrates is a man; therefore, Socrates is mortal” — eventually inspired Leibniz, the philosopher who imagined a “machine that could calculate thought.”
The Mathematics of Maybe
From Leibniz’s dream came binary logic — ones and zeros, the foundation of computing. Yet, Morley argued, life isn’t binary. “Logic lives at the extremes of true and false, but most of life happens in between.” Enter inductive reasoning, the world of probabilities, predictions, and generalizations — the domain where AI thrives. “Deductive logic gives you certainty,” he said. “Inductive logic gives you a headache —and progress.” He then dove — humorously — into the Monty Hall problem, birthday paradox, and
disease-testing dilemma, illustrating how humans instinctively misunderstand probability.
“We don’t ask what’s likely,” he smiled, “we ask what it feels like.” When you test positive for a rare disease with a 99% accurate test, he showed, the actual probability you’re ill might be just 9% — a humbling reminder that intuition is a poor statistician. The heart of his message came with a simple diagram of overlapping circles representing Bayes’ Rule. “This little formula,” he said, “drives nearly every AI model you’ve ever used — and can also make you a better person.” Bayesian reasoning, he explained, is about constantly updating beliefs based on new evidence. The one unforgivable sin? Refusing to change your mind. “If your prior belief is zero or one,” he warned, “no amount of evidence will change it. That’s not logic—that’s stupidity.”
When Probabilities Go Wrong
Morley illustrated how misapplied probabilities can destroy lives. In the O.J. Simpson trial, he noted, the defense misused statistics to imply that abusers rarely become murderers — confusing conditional probabilities and leading to a misleading conclusion. In the Birmingham Six case, six innocent men were imprisoned after juries mistook the probability of a positive test given innocence for the probability of innocence given a positive test. These examples, he said, show that “AI doesn’t make mistakes — people do, when they don’t understand the math.” He urged the audience to learn the difference between logical fallacies (flaws in argument structure) and cognitive biases (flaws in perception). “These are not bugs,” he noted, “they’re features of our brains.” Awareness of them, he said, is like installing a mental firewall.
Of Toast, Demons, and Soldiers
To prove that our brains are unreliable narrators, Morley flashed an optical illusion: two squares that appeared light and dark but were, in fact, the same color. Then came a sound clip from an old “Satanic panic” hoax: a rock song allegedly containing hidden messages when played backward. “You only hear the message after I tell you what to listen for,” he said. “That’s how bias works.” From there, he pivoted to psychology. Humans, he explained, have two cognitive modes— System 1, the fast, intuitive one that spots patterns (and ghosts in toast), and System 2, the slower, analytical one that questions them. “The trick,” he said, “is knowing when to switch it on.” He borrowed Julia Galef’s analogy of the soldier and the scout. The soldier defends beliefs; the scout seeks truth. The world, Morley said, needs fewer soldiers and more scouts. “Be the scout who asks, ‘Has it ever occurred to you might be wrong? ’— and then answer that question
honestly.”
Regaining Agency in the Age of AI
In his final act, Morley brought the discussion back to AI. Machine learning systems, he argued, are pure Bayesians — they update endlessly, probabilistically, imperfectly. Humans, however, resist change. “Technology isn’t destroying our critical thinking,” he said, “our unwillingness to think probabilistically is.” He reflected on the pandemic years — when people made life decisions based on COVID tests of dubious reliability — as a global case study in statistical misunderstanding. “If we can misread a simple test,” he asked, “how will we cope with probabilistic machines running our world?” Still, he was hopeful. “Everything is definitely maybe,” he concluded. “And that’s okay — as long as we keep looking up, keep looking for evidence, and keep our minds open.” He left the audience with one final, almost childlike line of wisdom, borrowed from Charlie Brown: “The secret of life,” Morley said softly, “is to keep looking up.” And with that, the mathematician who’d just turned Bayesian probability into philosophy received a thunderous applause — proof, perhaps, that critical thinking can still be entertaining.
Get your copy of the Autopilot Yes/No Report.
Please note – This report was created by almost exclusively using available AI-tools except for minor editorial tweaks and some limited lay-out changes.