The Open Society and Its Enemies

What Hacker News says

1. saubeidlDec 2025
   Again, I would recommend you familiarize yourself with the Paradox of Tolerance, written up by an Austrian philosopher after the horrors of WW2. Honestly, the entire book - The Open Society and its Enemies - might be a good read.

   We've been here before. We learned our lesson. We can't wait for Americans to catch up.

   thread
2. wizzwizz4Dec 2023
   I'm going to thoroughly trash-talk your work, but please keep in mind: just because you've produced something nigh-on-worthless, that doesn't mean you are worthless. On top of that, you've definitely got some skills, like writing clearly. (The ideas are nonsense, but I understand why they're nonsense, because you've presented your understanding very well indeed.)

   > Abstract

   Showing that all problems in NP are equivalent to problems in P is proving P = NP… but your abstract acts like that's a different thing. It's clear you don't even understand the question you're trying to answer. Most people would stop here.

   Most people would stop here, because they have found that there are certain people who you cannot convince of their own work's invalidity. Because replying means you will be sent papers to "proof-read" for the rest of the crank's life. I'm going to try putting exercises-for-the-reader in this criticism, and see if that helps, but I haven't got my hopes up.

   There is no "middle road". Either P = NP-Complete (the hardest problems in NP can be solved by algorithms in P), in which case P = NP; or P ∩ NP-Complete = ∅ (there is no P algorithm that can solve any NP-Complete problem). This is because all NP-Complete problems are equivalent with each other.

   Exercise: reduce the clique problem to 3-SAT.

   > Math IS Chemistry

   Sounds like you're vaguely thinking of https://en.wikipedia.org/wiki/Entropy_in_thermodynamics_and_... at a surface level, except all of the details are wrong. That diagram is nonsense.

   Exercise: calculate the minimum work required, in joules, to zero a 64-bit register with unknown contents at 20°C.

   > Introduction

   Equivalence relations have nothing to do with metrics. We are thinking clearly about the problem, and you aren't.

   Exercise: reach the Algorithm world of the Lean4 Natural Number Game: https://adam.math.hhu.de/

   > Words

   That's not polynomial time: that's quadratic time. Polynomial time is O(n^k) for some k. ≠ means "is not completely the same": completely different (for sets) can be written "A ∩ B = ∅".

   You're not even attempting to solve "P=NP?": you're tackling a different problem. There's no problem with that: the problem you've attempted should only take you about a year of proper study to learn how to solve, and I'd actually be interested to see you give a correct explanation. (You might find an answer more quickly than that.) The question is “Are there any quadratic-time algorithms in the class NP?”, and the correct explanation is about 3±2 sentences long.

   Exercise: look up the standard meanings of the words you've used. (Where you've said 'category', you mean 'class'; and 'stuff' should be 'physical object' or 'body'.)

   Exercise: derive Bayes' theorem.

   > 1. An argument of universal freedom […] because the universe would not exist.

   You have no reason to believe any of this: it's just baseless speculation.

   Exercise: read The Open Society and Its Enemies, by Karl Popper. 1947 book, published in two volumes (though there are combined editions). If it's not in your local library, you can borrow it from the Internet Archive: (1) https://archive.org/details/in.ernet.dli.2015.59272 and (2) https://archive.org/details/in.ernet.dli.2015.77661

   > What is the opposite of waste? Synthesis. A universe must generate stuff.

   This really doesn't follow from anything you've said. It doesn't even have a clear meaning.

   Exercise: practice deleting passages from your work. (If you can't bring yourself to do that, cutting-and-pasting them to a discard file is also acceptable, so long as you don't try to salvage them later.) Kill your darlings.

   > Why? […] arrow of time.

   Not entirely wrong! This whole tangent has nothing to do with P=NP, but it's kinda almost right!

   Exercise: find an example of a physical conservation law, and learn why people believe it's a conserved quantity.

   Exercise: find a physicist who thinks magnetic monopoles can exist, and a physicist who thinks they can't exist. Learn what their reasoning is. (Don't actually ask them directly: I mean find someone who's explained their reasoning already, and read what they wrote.) Think back on what you learned from The Open Society and Its Enemies: how does this relate to that?

   > Would it be possible for a universe to spontaneously invert encryption, simply to increase freedom? I would argue, yes.

   Exercise: find out what a "macrostate" is. Consider a discrete 52-card deck of cards, and a uniform shuffling procedure. Calculate the probability of the macrostate transitions (sorted → sorted, sorted → unsorted, unsorted → sorted, unsorted → unsorted). Consider the evolution of the system for the first five discrete time steps, starting at a sorted configuration.

   > Is a "hack" […] Kinetics.

   The mistakes you're making here should already be addressed by the earlier corrections.

   Exercise: explain why what you've written here is wrong.

   Exercise: learn what cybersecurity is.

   > 2. An argument that “solve” is equivalent to “check”

   First sentence is wrong. Third sentence (about “undecidable”) actually looks like computer science: I think your understanding here is correct.

   > Therefore, it is almost surely possible to convert “np” problems into “p” problems: generate candidate solutions from the problem itself, then check them.

   You can do this. However, you'll need to try exponentially-many candidate solutions, so the overall algorithm takes exponential time.

   Exercise: calculate the asymptotic time complexity of bubble sort, and deterministic bogosort. Compare the two.

   > For example, in 3-sat […] generate the candidate and check it.

   The interesting part about 3-sat – the bit that makes it NP-hard – is the fact that you can't, actually, do that. You can't iteratively go from a incorrect candidate solution to a less incorrect solution: the candidate solution might be wrong in a fundamental way, partly-correct only by accident, with no way to improve it.

   Exercise: understand that this is a metaphor for your paper. You can't make this paper correct by making iterative changes to it, because it's so fundamentally wrong that it can't ever be made correct.

   > existing SAT solvers

   … aren't in P. So you can't use them as part of a polynomial-time algorithm. Not even if you use a three-valued logic version of the SAT-solving algorithms: that doesn't change how fast they run, or really anything at all (since the algorithms, being two-valued logic algorithms, would just be ignoring one of the values of your three-valued logic).

   > 3-Sat Problem = (-1, 2, 3), (-1, -2, 3) => Solution to Check = {1: False, 2: Maybe, 3: True}

   This looks like the very beginnings of a heuristic (symbolic) 3-SAT solver. It's not polynomial-time, but it'd be slightly faster than the naïve brute-force approach. That's something to be proud of rediscovering!

   Exercise: implement a naïve 3-SAT solver (tries all the possibilities), then add a heuristic to make it a bit faster.

   > […] the solutions must share structure […] This means we can save incremental progress to speed up checks over time.

   I can see why you might think that (I used to think that as well), but it isn't true.

   Exercise: find a 3-SAT formula with two solutions that don't share structure. You probably changed your mind about what "structure" is half-way through, so now find a 3-SAT formula with two solutions that don't share your new definition of structure.

   > However, if we truly understand our problem, […] shrinks the category of solutions dramatically.

   Yes… if we just look at the first and last bit (the bit in the middle, about "Maybe", is wrong). And doing that is the hard part: we're trying to measure how long this process takes, when we're measuring the time complexity of an algorithm to solve a problem.

   Exercise: try to implement a 3-SAT solver that works the way you describe in the middle bit, and see why it doesn't work.

   > 3. An argument for universal search

   This whole section is completely wrong.

   Exercise: write an algorithm in P that verifies a solution to the Clique problem. (Finding a solution cannot be done in P, in the general case.)

   Exercise: write an algorithm in P that finds a solution to some special-cases of the Clique problem. Show that it cannot solve all cases in P.

   > Summary

   Pretty good summary of your arguments. If you figure out how to stop writing nonsense, you might make a decent academic.

   > Implications

   Exercise: research the Principle of Explosion.

   > References

   I see you've got a Category Theory reference. You do not understand Category Theory, and you will not be ready to learn about Category Theory for a very very long time, if ever. Most pure mathematicians never learn category theory. It's got nothing to do with your "category" (which is usually called "class").

   Other than that, this section is not bad!

   Exercise: learn a convention for including inline citations. (optional)

   Exercise: why are citations not usually URLs?

   (But yeah, overall, this paper is worthless. I spent entirely too long reviewing it. I can see you're passionate, but it wasn't worth me reading it, and I knew that from the start.)

   Show morethread
3. flanked-everglSep 2023
   I'm highly sceptical of the suggestion that if we just all take MDMA the increasingly fractured world will somehow re-unite. I think reading Karl Popper's open society and its enemies will be much more productive, as that will help you understand the reason for the division.

   It's as my old pappy used to say, you can't fix problems with more MDMA that wasn't created by a lack of MDMA.

   thread
4. seanhunterAug 2023
   This is an argument that seems reasonable but lacks important nuance and addressing it forms the center of "The Open Society and its Enemies" by Karl Popper, which I would recommend reading.

   Popper's argument is essentially that tolerance is essential to open society and therefore to maintain openness, society must ironically be intolerant of those who are intolerant themselves or would spread intolerance. Nazis were specifically on his mind, WW2 having just occurred, and Popper uses them as an example of people who use tolerance by others to spread intolerance and therefore reduce freedom overall. For society to defend freedom, it has to not permit that.

   Show morethread
5. tshaddoxApr 2023
   The dichotomy isn’t that we either laud the amount of bad things happening or stamp out freedom with tyrannical government. That’s precisely the ludicrous “worse is better” model I was criticizing.

   And I’m fairly aware of Popper. I largely share his views, particularly those in The Open Society and Its Enemies and The Logic of Scientific Discovery. I don’t think he would point to, for instance, high murder rates as an indicator of freedom in a society or a preponderance of pseudoscientific theories as an indicator of freedom in a scientific community.

   Show morethread
6. sukunrtDec 2022
   1. The art of doing science and engineering by Richard Hamming. This might be the best engineering book I've ever read. Hamming's breadth and depth are fascinating. His back of the envelope calculations are crazy accurate and early accounts of computing hilarious(real programmers write machine code). There is a great account of his discovery of error correcting codes and a great account of his failure to discover FFT. All in all it's a wonderful book on how to think about things in science and engineering.

   2. The open society and its enemies by Karl Popper. This is a great book on open society and liberal democracy. Broadly the book is a critic of historicist theories of Plato and Marx. There is also good philosophy on essentialism and the futility of getting closer to the truth by defining things more precisely.

   3. Printing Press as an agent of change by Elizabeth Eisenstein. This account of the impact of the printing press on Europe is very long but worth it. The parts on the scientific revolution are particularly good.

   Show morethread