Worse news: there are uncountably many undecidable problems. Computers have been a scam this whole time
"In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limits."
www.computerworld.com/article/4059...
www.computerworld.com/article/4059...
OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws
In a landmark study, OpenAI researchers reveal that large language models will always produce plausible but false outputs, even with perfect data, due to fundamental statistical and computational limi...
www.computerworld.com
September 21, 2025 at 8:46 PM
Worse news: there are uncountably many undecidable problems. Computers have been a scam this whole time
sorry, your True Name is undecidable
wouldn't it be cool to know your own minimum length description? you could quine yourself
October 13, 2025 at 5:58 AM
sorry, your True Name is undecidable
In one of Reddit’s transparency reports, they disclosed that ~94% of 31,000,000 user reports they received in 2021 were false, duplicate, & undecidable reports. 19 out of every 20, garbage.
First: Meta will not try to proactively detect "lesser" violations, instead relying on user reports. It is unclear (probably intentionally) exactly what this includes, but it definitely includes hate speech. Unfortunately, user reports are crap, have always been crap, and will always be crap.🧵 2/11
January 10, 2025 at 12:41 AM
In one of Reddit’s transparency reports, they disclosed that ~94% of 31,000,000 user reports they received in 2021 were false, duplicate, & undecidable reports. 19 out of every 20, garbage.
"being female" constricts women into being passive heterosexual subjects of the state, gender marked as eternal bodily subjugation, thus preventing liberation. "the female" is undecidable *because* it is social and shifting - you cannot have it w/o it being used as a police baton against women.
July 9, 2025 at 7:18 PM
"being female" constricts women into being passive heterosexual subjects of the state, gender marked as eternal bodily subjugation, thus preventing liberation. "the female" is undecidable *because* it is social and shifting - you cannot have it w/o it being used as a police baton against women.
this is very interesting (timestamped)
Mathematics foundations problem settled using AI: Medvedev logic is undecidable
YouTube video by Dr. Samuel Allen Alexander
youtu.be
September 21, 2026 at 11:17 PM
this is very interesting (timestamped)
December 6, 2024 at 2:12 AM
Classical billiards can compute.
With Isaac Ramos, we show that 2D billiard systems are Turing complete, implying the existence of undecidable trajectories in physically natural models from hard-sphere gases to celestial mechanics.
Determinism ≠ predictability. 🎱🧠 @upc.edu @ricardsole.bsky.social
With Isaac Ramos, we show that 2D billiard systems are Turing complete, implying the existence of undecidable trajectories in physically natural models from hard-sphere gases to celestial mechanics.
Determinism ≠ predictability. 🎱🧠 @upc.edu @ricardsole.bsky.social
December 19, 2025 at 8:16 PM
Classical billiards can compute.
With Isaac Ramos, we show that 2D billiard systems are Turing complete, implying the existence of undecidable trajectories in physically natural models from hard-sphere gases to celestial mechanics.
Determinism ≠ predictability. 🎱🧠 @upc.edu @ricardsole.bsky.social
With Isaac Ramos, we show that 2D billiard systems are Turing complete, implying the existence of undecidable trajectories in physically natural models from hard-sphere gases to celestial mechanics.
Determinism ≠ predictability. 🎱🧠 @upc.edu @ricardsole.bsky.social
muppet vore is just generally an unresolved and possibly undecidable problem, because of course you just run into topological limitations w/r/t what you do with the hand inside the muppet.
September 15, 2026 at 6:15 AM
muppet vore is just generally an unresolved and possibly undecidable problem, because of course you just run into topological limitations w/r/t what you do with the hand inside the muppet.
my theory of scientific progress and the rationally undecidable status of the question of pursuing it
August 2, 2026 at 11:19 AM
my theory of scientific progress and the rationally undecidable status of the question of pursuing it
If a mathematician is infallible does that mean they can decide the undecidable?
May 8, 2025 at 6:26 PM
If a mathematician is infallible does that mean they can decide the undecidable?
I got nerd sniped by Richard Stanley this week.
He answered a MathOverflow question about undecidable problems by suggesting that "people in the area believe" that it's undecidable whether or not a given polyomino, you can use it to tile a rectangle, as in the below picture.
He answered a MathOverflow question about undecidable problems by suggesting that "people in the area believe" that it's undecidable whether or not a given polyomino, you can use it to tile a rectangle, as in the below picture.
April 19, 2025 at 4:34 PM
I got nerd sniped by Richard Stanley this week.
He answered a MathOverflow question about undecidable problems by suggesting that "people in the area believe" that it's undecidable whether or not a given polyomino, you can use it to tile a rectangle, as in the below picture.
He answered a MathOverflow question about undecidable problems by suggesting that "people in the area believe" that it's undecidable whether or not a given polyomino, you can use it to tile a rectangle, as in the below picture.
It remains hilarious that parsing C++ is undecidable.
Parsing C++ is literally undecidable
Parsing, performance, and low-level programming.
blog.reverberate.org
April 23, 2026 at 8:42 PM
It remains hilarious that parsing C++ is undecidable.
New newsletter! This is an explainer on what "undecidability" means in computer science, why (from one perspective) it's not a big deal that the halting problem is undecidable, and why (from another perspective) it's a very, VERY big deal. buttondown.com/hillelwayne/...
What does "Undecidable" mean, anyway
An explainer for people who don't know computer science and are mildly curious
buttondown.com
May 28, 2025 at 7:25 PM
New newsletter! This is an explainer on what "undecidability" means in computer science, why (from one perspective) it's not a big deal that the halting problem is undecidable, and why (from another perspective) it's a very, VERY big deal. buttondown.com/hillelwayne/...
they should replace floats with exact reals, in which equality is an undecidable problem
August 19, 2025 at 4:29 PM
they should replace floats with exact reals, in which equality is an undecidable problem
Calculating π is a deterministic, primitive recursive recipe—no branching, guaranteed convergence. Proof discovery is undecidable (Church-Turing). Enumerating strings is intractable; what systems like AlphaProof do is heuristic search space pruning. Conflating the two is a category error.
September 25, 2026 at 3:36 AM
Calculating π is a deterministic, primitive recursive recipe—no branching, guaranteed convergence. Proof discovery is undecidable (Church-Turing). Enumerating strings is intractable; what systems like AlphaProof do is heuristic search space pruning. Conflating the two is a category error.
addicted to undecidable problems
addicted to bad computers
September 8, 2024 at 1:26 PM
addicted to undecidable problems
Chess is not undecidable, but in generalized form (i.e, on an nxn board) it is EXPTIME-complete. See: en.wikipedia.org/wiki/Game_co...
September 5, 2026 at 8:06 AM
Chess is not undecidable, but in generalized form (i.e, on an nxn board) it is EXPTIME-complete. See: en.wikipedia.org/wiki/Game_co...
Ruin a book’s title by replacing one word with “butthole.”
On Formally Undecidable Propositions of Principia Mathematica and Related Buttholes
On Formally Undecidable Propositions of Principia Mathematica and Related Buttholes
Ruin a book’s title by replacing one word with “butthole.”
I’ll do my own book.
Butthole Taste Texture
I’ll do my own book.
Butthole Taste Texture
Ruin a book’s title by replacing one word with “butthole”:
The Butthole of the Dragon
The Butthole of the Dragon
June 29, 2025 at 12:44 AM
Ruin a book’s title by replacing one word with “butthole.”
On Formally Undecidable Propositions of Principia Mathematica and Related Buttholes
On Formally Undecidable Propositions of Principia Mathematica and Related Buttholes
You’re equivocating on “property of the process.” Lack of guaranteed convergence isn't a property of consciousness; it's the nature of heuristic search over undecidable spaces, which humans and AI share. Leaping from search complexity to demanding "awareness" just smuggles in a ghost.
September 25, 2026 at 12:30 PM
You’re equivocating on “property of the process.” Lack of guaranteed convergence isn't a property of consciousness; it's the nature of heuristic search over undecidable spaces, which humans and AI share. Leaping from search complexity to demanding "awareness" just smuggles in a ghost.
a cursory wikipedia search tells me: totally undecidable, totally lost to history,
March 9, 2026 at 4:59 AM
a cursory wikipedia search tells me: totally undecidable, totally lost to history,
My reduction creates a deterministic system in dimension 6. The problem is decidable in dimension 1 even for nondeterministic systems. I don't know at which dimension things get undecidable.
September 25, 2026 at 2:25 PM
My reduction creates a deterministic system in dimension 6. The problem is decidable in dimension 1 even for nondeterministic systems. I don't know at which dimension things get undecidable.
the primary reason that this stuff is controversial is that there are people whose jobs we generally agree are useful who are not being paid for reasons unrelated to the undecidable question of whether a statistical abstract over some body of data can violate the copyright of the data it abstracts
one of the major problems with capitalism which AI exacerbates is that useful activity (search engines, news) is often entirely a side effect of useless activity (ads, ads), and we expect the distribution of that useful activity to remain constant even as the returns to related activities change
September 13, 2024 at 10:05 PM
the primary reason that this stuff is controversial is that there are people whose jobs we generally agree are useful who are not being paid for reasons unrelated to the undecidable question of whether a statistical abstract over some body of data can violate the copyright of the data it abstracts
Will just be posting what I think are interesting snippets.
"there are no clear frontiers that are separating the "AI-feasible" problems from the "AI-hard" problems (which certainly still exist, given that the difficulty level of problems are unbounded, and can even be undecidable). "
"there are no clear frontiers that are separating the "AI-feasible" problems from the "AI-hard" problems (which certainly still exist, given that the difficulty level of problems are unbounded, and can even be undecidable). "
September 8, 2026 at 8:42 PM
Will just be posting what I think are interesting snippets.
"there are no clear frontiers that are separating the "AI-feasible" problems from the "AI-hard" problems (which certainly still exist, given that the difficulty level of problems are unbounded, and can even be undecidable). "
"there are no clear frontiers that are separating the "AI-feasible" problems from the "AI-hard" problems (which certainly still exist, given that the difficulty level of problems are unbounded, and can even be undecidable). "