Lightnews — Scholar-powered news

LightNews

Gregor

@coexact.bsky.social

140 followers 340 following 820 posts

machine for turning coffee into other fluids

Posts Media Videos Starter Packs

Pinned

Gregor @coexact.bsky.social · Feb 4

Electric charge is an integer quantity in nature. It didn't have to be this way: We could imagine a world in which electrons and protons have charges that aren't precisely equal and opposed to each other. Chemistry would break! So where does this conspiracy come from? (it's topology) 🧵

Gregor @coexact.bsky.social · 2h

(this isn't entirely unrelated to the example of algebraic numbers in R. before schemes, people talked about 'general points' on varieties, meaning say some point 'without specific properties', ie not in any proper algebraic subset. The density of these directly translates into the density of eta)

Gregor @coexact.bsky.social · 2h

a single generic point is dense in Zariski, so is its complement..

Gregor @coexact.bsky.social · 4h

hmm well you're cursed I'm afraid, it seems to work within a minute usually: bsky.app/search?q=%40... (click on latest)

Gregor @coexact.bsky.social · 5h

wrong bot, try spelling it with a hyphen

Gregor @coexact.bsky.social · 1d

don't worry it's folklore

Gregor @coexact.bsky.social · 1d

I think it's often (exactly what I) mean to say

Gregor @coexact.bsky.social · 2d

@scgriffith.bsky.social

Gregor @coexact.bsky.social · 6d

it's vexing

Gregor @coexact.bsky.social · 6d

event as a cause, but for a logical framework that has a more digestible minimal set of assumptions yet implies the things we want to explain.

Gregor @coexact.bsky.social · 6d

hmm so idk what the math philosophers think abt this but imo the sense in which the PH explains the entropy gradient is the sense in which we have mathematical theories to explain purely mathematical facts, in which domain there also is no real causation. We're not looking for a temporally prior

Gregor @coexact.bsky.social · 6d

We were just talking about this over here!

Rochelle @uberwensch.bsky.social · 8d

If you're talking about the idea of the Past Hypothesis, I do think its a mistake to look to initial (or early conditions) to "explain" the arrow of time, because stipulating those facts is tantamount to stipulating all facts. So it's like saying "things equilibrate because they do."

Gregor @coexact.bsky.social · 7d

and ofc it does bc how else to connect to observation? "The fact that we can define coherent clock settings [..] to give invariant light speed is a highly non-trivial attribute of the physical world, one that has been ascertained empirically."

Gregor @coexact.bsky.social · 7d

well they suck ><

but fr I don't get their issue. they seem to just dislike clocks bc they're too complicated. But for one that's not true, the design of the clock is irrelevant, and for another their own explanation at the start uses clock synchronization

Gregor @coexact.bsky.social · 7d

u mean talking at Rochelle I assume 🙂‍↕️

Gregor @coexact.bsky.social · 7d

tragic that this wasn't translated more literally as *hole inspection*

Gregor @coexact.bsky.social · 7d

"this is not a VEGAN BURGER according to EU regulation 1308/2013" and make VEGAN BURGER very large on the packaging

Gregor @coexact.bsky.social · 7d

this is because a rising sea is not étale 😶

Gregor @coexact.bsky.social · 7d

(and they might not be: I think if eternal inflation is true for instance there is no upper bound to global entropy so any state is 'far from equilibrium')

Gregor @coexact.bsky.social · 7d

On the other hand, for any disequilibrium state you pick you should generically expect a global entropy increase in both directions of evolution as I pointed out above. It is not hard to find a disequilibrium state even though they might also be thin by some measure

Gregor @coexact.bsky.social · 7d

I think it is extremely unsurprising. to construct a state in which one part of an interacting system equilibrates while another disequilibrates is some kind of impossibly hard sudoku puzzle. To find yourself in such a world might be reason to believe in God.

Gregor @coexact.bsky.social · 7d

(by past evolution I mean that I expect entropy to increase towards the past in much the same way as towards the future. Also there are many difficulties in making this story super precise but I don't see any real reason not to believe it in principle)

Gregor @coexact.bsky.social · 7d

given an arbitrary state, if I randomly perturb it a tiny amount, its future (and past) evolution will with overwhelming likelihood follow the laws of thermodynamics, and this is not true of any other attempts at macroscopic laws I could write down!

Gregor @coexact.bsky.social · 7d

I'm not sure I buy that the first sentence is enough to justify the second. The laws of thermo are coarse-grainings, it's just that they are statistical in nature (which we already knew), applying only to almost all states; producing a counterexample doesn't disprove anything.

Gregor @coexact.bsky.social · 8d

no, the whole lightcone (minus the vertex) is one orbit for O(d,1). And a null vector is stretched by a boost in a way that directly corresponds to redshift!

Gregor @coexact.bsky.social · 8d

It's an observation we can make that when transporting each clock sufficiently slowly according to its own time, they stay synchronized (within some accuracy) when meeting back up. This is observer-independent, though each clock showing the 'right' time isn't