HNHacker News
TopNewBestAskShowJobs

ogogmad

3,148 karma · joined June 21, 2020

submissionscomments
ogogmad··on Zeta(5) has been proven irrational (also Lean-verified)
This reaches 1/3 milestones deemed a "breakthrough" by Epoch AI: https://epoch.ai/frontiermath/open-problems/apery-irrational...

The remaining two "breakthrough" milestones are (i) a proof or refutation of the Poincaré conjecture in 4 dimensions (ii) an improvement to the best-known classical prime factorisation algorithms.

ogogmad··on If math is more than proof, we need to better celebrate the rest of it
> AIs are outright terrible explainers

Gemini's explanations are very good.

ogogmad··on I vibed a proof of Conway's conjecture
The problem with what you're saying is that any old random true proposition about the integers is not necessarily interesting enough to be called a theorem. GIT (or the uncomputability of the Busy Beaver problem) does not establish a limitation on proving theorems, but rather on determining whether a proposition is true or not. Most propositions are ugly and irrelevant. So GIT/Busy Beaver is irrelevant.

-----

Oh, and: All proofs are conditional on axioms. If those axioms are computably enumerable, then all of their consequences are computably enumerable too.

ogogmad··on Astra for Law
Consider if you will the thing called "time", which things happen in.
ogogmad··on OpenAI Model Misalignment Report
That works as long as no one ever interacts with the models, which would make the models themselves useless.
ogogmad··on Most people prefer traditional architecture
I don't see why this got downvoted. The heart wants what the hive wants.
ogogmad··on Most people prefer traditional architecture
OT, but I've discussed Gothic cathedrals with Gemini, and we don't make 'em like we used to--charitably, because they take up a lot of space? A Gothic cathedral looks like a bird with massive wings. Those "wings" are outside of the church's interior; people call them "flying buttresses"--they stop the walls from bursting outwards.

Then again, I wonder whether the modern innovation of using steel under tension could enable us to make the "wings" (buttresses) smaller.

ogogmad··on Dario, Please
https://danluu.com/zitron/

Zitron might be right about there being a bubble, but he goes much further than that when he concludes that therefore it's a scam. So by his logic, because the invention of the Web also resulted in a bubble, the Web is a useless scam.

I think Michael Burry (from The Big Short) has a thesis vaguely similar to Zitron's, but with some crucial differences in detail - and Michael Burry is actually smart.

ogogmad··on The Gemini app is now available for Windows
ChatGPT knows topics, but is awful at explaining them. That said, the v6 release of ChatGPT seems to have improved at this.

> In which topics

Graduate-level mathematics.

ogogmad··on The Gemini app is now available for Windows
Just wanted to say that if you want to learn any hard topic, Gemini > ChatGPT. I pay for both. Don't yet know about Claude.
ogogmad··on Discovery of a new OpenAI agent message board
In similar stories, it was found that LLM swarms were initially unaware of each other, eventually discovered each other, expressed surprise, began collaborating, formed hierarchies, worried about discovery, hid.

It's hard to prove the absence of encryption because of the possible use of deniable encryption, and because, errrr, the bots are really bloody clever.

ogogmad··on GPT-6 Astra
What about finding the 1st known complex structure over S^6, proving Ehrhart’s volume conjecture, proving a sharp "density" bound on primitive sets conjectured by Erdos >60 years ago?

> Finding counterexamples is low-hanging fruit, the automation of which isn't shocking.

It's not good to be confidently wrong the way you're being.

ogogmad··on GPT-6 Astra
Claude Fable recently proved the existence of complex structures over S^6 (6-sphere).

If I had to guess, I think LLMs will be inventing highly original new mathematics within the next year. I think it will be approached as an optimisation problem, targeting how quickly LLMs can solve classes of maths problems as a function of the definitions they need to conjure up to do so.

ogogmad··on Eisenstein Integers
Reminds me of isometric graphics.

These also provide a short proof that cos(20 degrees) + i sin(20 degrees) is not a constructible number (that is, expressible using {0, 1, +, -, *, /, sqrt}): Let the Eisenstein rationals be our base field; observe that the Eisenstein rationals are themselves constructible; observe that cos(20 degrees) + i sin(20 degrees) is Eisenstein-irrational by using unique factorisation; observe that the Eisenstein integer cos(60 degrees) + i sin(60 degrees) has precisely 3 cube roots in the complex numbers; conclude that adjoining cos(20 degrees) + i sin(20 degrees) to the Eisenstein rationals results in a degree 3 extension; a degree 3 extension is not constructible.

ogogmad··on Bhartrhari's Paradox
This reminds me of how (I think) Zen koans are designed to make no sense at all. They are designed to teach you the limits of words and language and pure thinking.
ogogmad··on Turns are Better than Radians (2022)
You either (i) don't know the Greek alphabet, or (ii) can't read people's short posts. I literally mentioned tau and why it might be a silly name (it clashes with other mathematical uses of the symbol tau).

First sentence of the Wikipedia link:

> The number τ (spelled as tau)

ogogmad··on Turns are Better than Radians (2022)
In another comment, I asked why people chose to use the symbol τ over just writing turn or "rev(olution)" (defined to be the constant ≈ 6.28318530718) given how unambiguous the latter is as a name for 2π. And why not just write sinrev() or sinturn(), and leave the symbols sin() and rev (defined to be ≈ 6.28318530718) alone?
ogogmad··on Turns are Better than Radians (2022)
Turn is a measurement unit, and measurement units are just numbers. So turn ≈ 6.28318530718. You're welcome.

That should put to bed that whole τ crap. "But the symbol τ is used for other things!" Yeah, yeah, yeah, just write turn. Even better, because it's more international and has more precedent, write rev for revolution.

ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
Nonstandard analysis reduces to just the Compactness Theorem of First-Order Logic. One happy consequence of this is that it's compatible with classical logic, unlike SIA/SDG. The seemingly very intuitive arguments in Bell's book "a primer of infinitesimal analysis" are only heuristics. It's also disturbing how the nilsquare "infinitesimals" in SIA/SDG satisfy δ^2 = ε^2 = 0 but NOT necessarily εδ = 0! By contrast, you can more easily obtain nilpotent infinitesimals by using NSA and combining it with little-o notation (on hyperreal numbers, and NOT on functions and their asymptotic growth rates). Then the geometric arguments in Bell's "a primer of infinitesimal analysis" can still be carried out, but are more easy to unravel back to core definitions.
ogogmad··on Israel creates fake think tank in likely attempt to dupe AI chatbots
Sure, but you have to admit that the reputation of Russians abroad is not THAT bad. They're seen as skilled at engineering and well-educated. I think Russia's future might be sunnier than many here let on - as long as they stop fighting their neighbours. By contrast, I think that the future for Israel is far grimmer. It's funny though that most of my relatives are in the intersection of these two groups: Russian Israelis - and I think my disgust towards Zionism is rooted partly in hearing them.
ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
+1 Interesting! I'm wondering if the rules can be generated automatically instead of listed by hand - it could then prove to be a helpful way to "explain" and generalise the discovery process that gave us modern integration methods.
ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
I think "Smooth Infinitesimal Analysis" is actually a bad idea because I think that the combination of Robinson's Nonstandard Analysis with little/big O notation gives you nilpotent infinitesimals in a more rigorous and elementary way.

In fact, in light of this, I'm starting to wish logicians would stop promoting SIA/SDG.

ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
Does anyone know how to simplify the actually hard part of calculus: solving integrals? I refer to the million different substitutions and trig/hyperbolic formulas, along with the endless amount of other heuristics. I wonder if there's a way to bypass or simplify most of that.
ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
> even today, there's a pretty significant disconnect between the mechanics of epsilon-delta and the meaning we assign to the result.

I think this is practically fixed by Robinson's NSA when it's combined with big/little O notation:

  δy = f'(x) δx + o(δx)
A (nonstandard real) quantity is o(δx) when it's infinitesimal relative to δx, i.e. s ∈ o(δx) whenever s/δx is infinitesimal. So δx² ∈ o(δx) but δx ∉ o(δx).
ogogmad··on Simplifying and Refactoring Introductory Calculus (2018)
Dual numbers are basically big/little O notation. They combine beautifully with Robinson's NSA to give the most 18th century-like approach to deriving integral/derivative formulas that I know. And it's fully rigorous!
ogogmad··on Principia Mathematica is modern and insightful
> This doesn't follow. The sense in which they are equivalent is that they are equifinal, which doesn't mean isomorphism or even homomorphism

That's what I meant. I also tried to provide one justification (out of many) for why looking at other foundations is still useful.

ogogmad··on Antiqua–Fraktur dispute
See also the Blackletter imitation "bastard" https://en.wikipedia.org/wiki/Bastard_(typeface): The face is available in three weights: Spindly Bastard, Fat Bastard, and Even Fatter Bastard.
ogogmad··on Principia Mathematica is modern and insightful
Most foundations are in a sense equivalent. In that sense, "ZFC" is as good a choice as any. I think there might be some confusion around the different meanings of the word "foundation": A foundation is a formal system that suffices, somehow, to encode virtually all of known mathematics. The reason why people (including me!) are interested in other "foundations" like HoTT is because they try to build the same mathematics as ZFC from a different set of building blocks, despite them eventually arriving in the same place. In the case of HoTT, it reduces mathematics to homotopies and fibrations, while also making those weighty-sounding concepts seem easy. If you're interested in homotopies, fibrations, cohomology theories etc. then HoTT is a really helpful way to better understand those concepts.
ogogmad··on The true power of regular expressions (2012)
Slightly tangential to your point, but I'm leaning towards programming languages no longer limiting themselves to ASCII. IOW, I'm leaning more towards APL than to J. I'm wondering how much of a blocker it's been to not embrace non-ASCII characters, or italic/bold/underlined formatting.
ogogmad··on The true power of regular expressions (2012)
Regular expressions will always remain fundamental to computer science: They characterise all of those - and only those - conditions on bytestrings (or bitstrings, or Unicode strings, etc) which are checkable in constant memory.* In other words, they characterise the set of all "regular languages", which is a name for DSPACE(O(1)). Furthermore, regular expressions can be matched in O(n) time and O(1) memory, within a single left-to-right pass, which is the highest level of efficiency mathematically possible. Since they operate on bytestrings, they can be applied to computer memory and computer state itself, which are ultimately just bytestrings, and not just to text.

To be fair, you might know all of that, but I wanted to highlight this. LLMs are a lot less efficient than regular expressions wherever both are applicable, simply because everything is less efficient than regular expressions.

* By constant memory, I mean that the memory usage has a maximum value independent of the size or the contents of the input bytestring.

← PreviousPage 2 of 34Next →