frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Debunking Classical Pseudo-Paradoxes of Logic

2•kokhanserhii•1h ago
You know, many textbooks on mathematical logic look more like textbooks on scholasticism and metaphysics. On one hand, they promise to show you the strictest foundation of thought — and on the other, they tell you: "Well, but if things get even slightly more complex than arithmetic, contradictions arise, proofs become impossible, incompleteness shows up, here are your paradoxes." If a system is just a bit more complex than arithmetic, it's already somehow defective — and to explain all of this, you need an even more complex system, which is itself, naturally, also "defective," incomplete, contradictory… In the end, any ordinary student drowns in all this like in some set of riddles the human mind is simply not equipped to solve. So what comes next? Essentially nothing — you're calmly taught logic, told to use it, everything is fine, it shows things clearly and precisely, the most complex algorithms work, calculations come out right. So what was the point of that digression in the course about contradictions and paradoxes? In this article, I try to briefly and precisely explain to the student — at a basic level — and resolve these non-constructive pseudo-problems. This applies not only to students. Specialists who were put through this scholastic mis-education in their university years would also do well to spend 10–15 minutes to quickly and simply sort this out — and regain confidence in their own logic. After all, you've noticed throughout your life that logic is sound and works just fine, haven't you? https://zenodo.org/records/19220604

Comments

PaulHoule•1h ago
https://en.wikipedia.org/wiki/G%C3%B6del%27s_incompleteness_...

is no joke. I've heavily studied applications of logic to computing from the golden age of AI and I can say it is thoroughly depressing to see how much "it doesn't work" which you try to do things with logic.

kokhanserhii•1h ago
It seems my article also refutes Gödel, because his proof is based on the liar paradox.
PaulHoule•55m ago
For logic to really go on wheels you need decision procedures that are decidable and for many of the logical systems you would want to use for real-life applications like engineering and bank regulation you want FOL + arithmetic which very much has problems. It's why we get stuck with things like OWL that seem overcomplicated and underpowered at the same time.

Don't get me started on how real commonsense reasoning requires partitioned knowledge bases, modal logic, temporal logic, social logic (Jim thinks that Mary believes ...), all sorts of complications for which a complete and consistent decision procedure is just unthinkable. I think you could put together a logic system for families of problems in certain domains and accomplish a lot but there is no general logic for real life and efforts like Cyc have gotten stuck in the mud or gone down in flames.

We Rebuilt Zeta from the Training Data Up

https://zed.dev/blog/zeta2
1•meetpateltech•2m ago•0 comments

Sanders and AOC unveil data center moratorium bill

https://www.axios.com/2026/03/25/sanders-aoc-data-center-moratorium-bill
1•blinding-streak•2m ago•0 comments

AI Might Be Our Best Shot at Taking Back the Open Web

https://www.techdirt.com/2026/03/25/ai-might-be-our-best-shot-at-taking-back-the-open-web/
1•speckx•2m ago•0 comments

What Will It Take to Build the Largest Data Center?

https://spectrum.ieee.org/5gw-data-center
1•Brajeshwar•2m ago•0 comments

Meta and Google found liable in social media addiction trial

https://www.bbc.co.uk/news/articles/c747x7gz249o
1•ColinWright•4m ago•1 comments

Musk lawyers try to bar judge over LinkedIn 'like' cheering legal defeat

https://www.ft.com/content/c53cfaa8-8db1-49b1-b877-3ac7b2095966
1•1vuio0pswjnm7•6m ago•0 comments

Show HN: Sift – local, de-noised GitHub notifications

https://github.com/vignesh07/sift
1•eigen-vector•8m ago•0 comments

Show HN: PromptPrivacy – Open-source tool to track how AI models use your data

https://prompt-privacy.vercel.app/
1•Sonofg0tham•8m ago•0 comments

Metasystemic

https://metasystemic.xyz/stuff.html
1•gdss•8m ago•0 comments

Show HN: TexoCAD – Describe a part in words, get actual editable CAD

1•torayeff•8m ago•1 comments

RAG to riches: synthetic data for training RAG agents

https://cgft.io/blog/rag-to-riches/
1•kumama•9m ago•0 comments

Data Is Now the Front Line of Warfare

https://www.csis.org/analysis/data-now-front-line-warfare
1•kdavis•10m ago•1 comments

Newsom calls Musk "one of the great disappointments" of this era

https://www.axios.com/2026/03/25/gavin-newsom-elon-musk-tesla-china
2•1vuio0pswjnm7•10m ago•0 comments

Show HN: Vellyr – small JavaScript non-lazy signals library

https://github.com/Bloomca/vellyr
1•bloomca•11m ago•0 comments

Meta hit with $375M verdict in New Mexico child safety case

https://www.politico.com/news/2026/03/24/meta-hit-with-375m-verdict-in-new-mexico-child-safety-ca...
1•1vuio0pswjnm7•11m ago•1 comments

The Bee That Everyone Wants to Save

https://naturalist.bearblog.dev/the-bee-that-everyone-wants-to-save/
2•nivethan•12m ago•0 comments

Average gas price poised to hit $4 a gallon

https://www.consumeraffairs.com/news/average-gas-price-poised-to-hit-4-a-gallon-032426.html
1•geox•13m ago•0 comments

Porting Go's io package to C

https://antonz.org/porting-go-io/
1•ashvardanian•14m ago•0 comments

A Cat Left Paw Prints on the Pages of This Medieval Manuscript

https://www.smithsonianmag.com/smart-news/a-cat-left-paw-prints-on-the-pages-of-this-medieval-man...
1•jruohonen•14m ago•0 comments

Meta and YouTube Lose Landmark Social Media Trial

https://www.wsj.com/tech/meta-and-youtube-lose-landmark-social-media-trial-33e4c5cb
1•jaredwiener•15m ago•1 comments

Google Voice suddenly requires an ID to text

https://old.reddit.com/r/Googlevoice/comments/1s3ha6h/suddenly_asks_for_my_id_to_send_a_text/
1•cft•16m ago•0 comments

ARC-AGI-3 Leaderboard

https://arcprize.org/leaderboard
1•yusufozkan•16m ago•0 comments

The AI Depression

http://charleshughsmith.blogspot.com/2026/03/the-ai-depression.html
1•speckx•17m ago•0 comments

Meta and YouTube Found Negligent in Landmark Social Media Addiction Case

https://www.nytimes.com/2026/03/25/technology/social-media-trial-verdict.html
25•mrjaeger•17m ago•0 comments

Should I Iron This Shirt?

https://weirdtools.xyz/should-i-iron-this-shirt/
1•kadirmalak•18m ago•0 comments

Google is expanding to neutral atom quantum computing

https://blog.google/innovation-and-ai/technology/research/neutral-atom-quantum-computers/
2•ernesto95•19m ago•0 comments

Long-Form Comes to Bridgy Fed

https://blog.anew.social/long-form-comes-to-bridgy-fed/
1•Kye•19m ago•0 comments

Show HN: Τ³-Bench is out – can agents handle complex docs and live calls?

3•victorbarres•21m ago•0 comments

Apple Can Create Smaller On-Device AI Models from Google's Gemini

https://www.macrumors.com/2026/03/25/apple-google-gemini-distill-models/
4•thm•22m ago•1 comments

Show HN: AI Chat that dynamically assembles UI over time

https://beta.kailaido.com/
1•terryr518•22m ago•0 comments