
Oct 17, 2026 · 1h 12m
Listen from 1:08:11
Listen at 1:08:11
David Deutsch argues that knowledge grows through criticism
Father Of Quantum Computing: How To Think Like A Genius - David Deutsch - #1164
The conversation connects fallibility and open-ended knowledge to debates about AI, climate change, education, and how people should live.
- 1Knowledge advances through conjecture and criticism, not certainty, authority, or the passive accumulation of information.
- 2Deutsch sees optimism as confidence that problems reflect missing knowledge rather than permanent limits or guaranteed success.
- 3Current language models may be impressive without being genuinely creative, because functionality alone does not establish new knowledge creation.
Don't miss
Deutsch distinguishes the impressive functionality of current language models from genuine creativity, which he associates with creating new knowledge.
The brief
David Deutsch frames human knowledge as conjectural and fallible: progress comes from criticism and better explanations, not certainty or deference to authority.
The Fermi paradox illustrates how easily people invent explanations, while Deutsch defines optimism as treating failure as evidence of missing knowledge rather than an unbreakable barrier.
Deutsch questions whether large language models can become genuinely creative by scaling alone, distinguishing impressive functionality from the creation of new knowledge.
His education critique rejects the idea that students are empty containers, favoring active learners who develop and test ideas rather than merely receive information.
The conversation ends with a similarly pluralist view of life: discipline should reduce unnecessary suffering, and no universal formula can fit radically different people.
What was said on this episode
39 statements · 16 positive · 17 negative · 6 neutral
Knowledge is created through one fundamental process across evolution, science, mathematics, and learning.
“there is only one way of creating knowledge”
Listen at 2:39
People should prioritize correcting errors rather than seeking certainty.
“we should not be looking for certainty. We should be looking to correct errors.”
Listen at 6:21
Subjective certainty does not reliably indicate truth.
“the subjective sensation of being certain is no guide to the truth. It's no guide at all.”
Listen at 7:49
Ideas should be judged by explanatory quality rather than authority or confidence.
“the only way to judge ideas is not by someone's authority or by one's own authority or by how sure one is, but only by the criterion of whether you have a good explanation”
Listen at 9:06
People should use theories about institutions and systems for practical decisions.
“The rational thing to do, as I said, is to form a theory.”
Listen at 10:57
Reliability should be assessed through explanatory theories rather than reputation alone.
“I don't go by their reputation. I go by that theory.”
Listen at 12:30
People should not aspire to certainty.
“No, we should aspire not to have any.”
Listen at 13:21
People should not treat belief itself as an aspiration.
“I don't think one should even aspire to believe anything.”
Listen at 13:56
Twitter or X is well suited to sharing conjectures.
“Twitter or X is the ideal medium for conjecture”
Listen at 17:33
Deutsch’s Fermi-problem conjectures are probably false despite his stated certainty.
“I am sure, certain, you know, which I shouldn't be, that they are not true.”
Listen at 19:23
Humanity does not know whether extraterrestrial life exists.
“We don't know whether we are alone in the universe. We might be or we might not be.”
Listen at 22:39
Failures result from lacking knowledge needed to achieve the desired outcome.
“whenever something goes wrong, it's because we had not yet created the knowledge to make it go right.”
Listen at 24:21
Improving from a bad situation to a good one is always possible.
“Going from bad to good is always possible.”
Listen at 25:29
No problem should be regarded as beyond reason’s potential reach.
“you do not think this is insoluble. You do not think this is how far reason can reach. And from now on, reason doesn't have any access. That's always false.”
Listen at 28:20
No problem should be regarded as permanently beyond the reach of reason.
“you do not think this is how far reason can reach. And from now on, reason doesn't have any access. That's always false.”
Listen at 28:24
Fear of catastrophic AI risk is not currently justified.
“So I think it's not a justified fear”
Listen at 34:59
Turning away from nuclear power caused humanity’s worst environmental effect.
“The worst effect on the environment that our species has ever had was the turn away from nuclear power.”
Listen at 35:52
Rejecting artificial intelligence because of unfounded fears would cause a major catastrophe.
“if we turn away from artificial intelligence because of a similar fiction, will it be? No, I was going to say it might well be a bigger catastrophe. I think I was a bit tempted to say something dramatic. No, I think it would be a big catastrophe, but not as big.”
Listen at 36:04
Current large language models lack a component capable of creativity.
“the LLMs do not have a module in them capable of creativity.”
Listen at 38:06
Current LLM-based AI systems have shown no sign of artificial general intelligence.
“the present LLM-based AIs, not only have they not shown any sign of AGI”
Listen at 38:40
Researchers do not know how to program grandmaster-like chess insight or AGI.
“We do not know how to write the grandmaster type of chess program, let alone an AGI.”
Listen at 43:42
Creativity is the ability to create new knowledge.
“Creativity, I think, is not that impressive a thing, just creativity by itself. I use that word to mean the ability to create new knowledge.”
Listen at 44:22
Humans uniquely possess explanatory creativity among existing species.
“The thing that humans have and that no existing other species has, although there were species in the past that had this, is explanatory creativity.”
Listen at 45:06
Humans can revise their goals and criteria rather than optimize only fixed objectives.
“Humans can change the axioms to better axioms, and humans are not just meeting a fixed criterion”
Listen at 47:25
Modern education is based on an incorrect theory of knowledge.
“our society's present... attitude to education, both for children and in universities and for everybody in their own life, is based on a wrong theory.”
Listen at 48:19
Educational incentives shift effort toward measurable criteria and away from important unmeasured abilities.
“The incentive is to spend more and more time on meeting the criterion and less and less time on things that can't be measured, which are the important things.”
Listen at 50:50
Important educational outcomes are generally not directly testable.
“everything that is important in education is of that type and not of the testable type.”
Listen at 51:35
Courses and exams are poor proxies for the abilities education actually seeks.
“The course and the exam are designed as proxies for the thing that you actually want. And they are very poor proxies.”
Listen at 52:26
Internet technologies could enable eliminating compulsory education, especially for children.
“we do now, with the internet, we do have the means of... completely upending the system and not having any compulsory education, certainly for children.”
Listen at 54:06
Internet technologies could enable eliminating compulsory education, including for children.
“we do have the means of... completely upending the system and not having any compulsory education, certainly for children.”
Listen at 54:09
Society should eliminate compulsory subjects and compulsory education.
“it would have been better, in my opinion, not to have compulsory subjects and not to have compulsory education of any kind.”
Listen at 55:19
Learning should involve pursuing difficult activities one values deeply.
“All learning should be like that.”
Listen at 58:19
A rational society would prevent students from cramming and immediately forgetting coursework.
“In a rational society, that would never happen.”
Listen at 1:00:20
Nearly every everyday situation can be improved somewhat through knowledge.
“There are no situations or almost no situations in everyday life that cannot be improved a little.”
Listen at 1:01:39
Reducing disliked obligations is always beneficial.
“Reducing the thing you don't like is always good.”
Listen at 1:03:20
No universal formula for a good life applies equally to people.
“I don't think there's a formula which applies to everybody. In fact, I don't think there's a formula that applies to any two people.”
Listen at 1:03:43
No single formula for a good life applies to everyone.
“I don't think there's a formula which applies to everybody.”
Listen at 1:03:43
Society often overlooks effort and creativity in socially disapproved activities such as video games.
“They will be blind to the effort and creativity that's going into it.”
Listen at 1:08:11
One of Deutsch’s books may be released within the next year.
“But in the next year, fingers crossed, yes. One of them may drop.”
Listen at 1:11:23
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.
Books & mentions
Some links are affiliate links — PodLume may earn a commission if you buy.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.

