iiinova — where your digital self keeps evolving

Molossus Spondee @molossus-spondee.bsky.social · 02/10/2026

Gay rights is bourgeois and that's a good thing. Gay rights is one of the last instances of the bourgeoisie being a progressive force. Abolishing the family is the end of liberalism and the beginning of Marxism.

000

Molossus Spondee @molossus-spondee.bsky.social · 30/09/2026

I don't think people actually engage with the specifics of nationalist movements in the imperial core. People speak of Indigenous nationalism, Black nationalism and other movements as abstractions. Precisely what organizations are we talking about?

101

Molossus Spondee @molossus-spondee.bsky.social · 29/09/2026

It's something silly like I want a map between the Lipschitz free space of a complete ordered family of equivalences and the Berkovich polydisk over 𝔽₂((t))

000

Molossus Spondee @molossus-spondee.bsky.social · 29/09/2026

Huh, I think I really just want to work in the sheaf topos over the Berkovich unit polydisk Sh(ℳ(𝔽₂((t))⟨x⟩)). The topos approach makes working with tree structures like magmas and abstract syntax trees a lot easier. The trouble is that it's hard to mechanize.

000

Molossus Spondee @molossus-spondee.bsky.social · 22/09/2026

As a transgender woman I'm pretty opposed to democracy, human rights and horizontalism. When people democratically vote to kill you, it makes you lose trust in democracy pretty quickly.

010

Molossus Spondee @molossus-spondee.bsky.social · 18/09/2026

I think the reality is that birth control alone is a big reason for the imperialists to hate trans women. :/

000

Molossus Spondee @molossus-spondee.bsky.social · 11/09/2026

I still don't have it right but I think you want something stupid simple like K = 𝔽₂((T)) R = K{X} ⟦ - ⟧ : X → R ⟦ E ⟧ · (E' Tⁿ) = [E =ₙ E'] T(n) = R{x₁, ..., xₙ} Ω(n) = Λ (T(n)ⁿ)* ∇ : T(n) → Ω¹(n) ∇ u · (v dxₖ) = T⁻¹([xₖ := T xₖ]u + u) · v Still seems ick

101

Molossus Spondee @molossus-spondee.bsky.social · 11/09/2026

Huh so mapping into rigid analytic spaces is actually very easy, you just want a transport functor out of the path groupoid into square matrices of convergent series. K = 𝔽₂((T)) tr: Path(X) → B(K{X}) tr(n, p: E =ₙ E') = Tⁿ (E ⊗ E') Something like that.

000

Molossus Spondee @molossus-spondee.bsky.social · 08/09/2026

youtu.be/bxQgqeF4Cqw I feel like Trump's push against "DEI" is in part because the big AI and ATS companies didn't want lawsuits on hiring discrimination. It's so stupid.

youtu.be

Yes, You ARE Getting Auto-Rejected

YouTube video by Urgently Firing

011

Molossus Spondee @molossus-spondee.bsky.social · 04/09/2026

I have decided that Marx was wrong. People have to fight for their own freedom before they can fight for others. It's not possible for the imperial core to fight for the periphery until the workers fight for their own freedom. The imperial core just has this weird charity ethic.

120

Molossus Spondee @molossus-spondee.bsky.social · 03/09/2026

It's stupid how bigoted leftists attack queer people for being assimilationists and for being lumpen apart from the class struggle. The demand to assimilate into the proletariat is a proletarian demand.

100

Molossus Spondee @molossus-spondee.bsky.social · 02/09/2026

For working with domain theory and bisected ultrametric spaces, I think I want something like synthetic differential geometry but for affinoid spaces (so probably a sheaf topos over 𝔽₂((t))-fp-AffSp). I just don't understand the maths involved.

000

Molossus Spondee @molossus-spondee.bsky.social · 01/09/2026

Marxism is not about whether the proles are good people or not. The Marxist understanding of history is that the workers will be compelled to rebel as exploitation intensifies. This in no way means that the working class will not have whatever ignorant and petty faults.

110

Molossus Spondee @molossus-spondee.bsky.social · 29/08/2026

I think you might want a variation on the exterior algebra something like the following for characteristic 2 Basically, you want a diagonal R[eₖ, δₘₖ | m < n, k < n]/((eₘ eₖ δₘₖ)(δₖₖ - 1))

100

Molossus Spondee @molossus-spondee.bsky.social · 28/08/2026

People get really weird about marginalized people being sexist. Revolutionaries have had to navigate literal blood feuds between clans that have been killing each other for generations. Some bullshit petty superstitions are not so different.

100

Molossus Spondee @molossus-spondee.bsky.social · 26/08/2026

The thing about job search advice is that nobody is going to admit they got a job through nepotism. They're going to refer to a good portfolio or obliquely to the importance of networking.

010

Molossus Spondee @molossus-spondee.bsky.social · 25/08/2026

I think I probably just want some modules over rings along the line of S = R ⊕ ⨁ n. R[h]/(hⁿ⁺² - h) S = R ⊕ ⨁ n. R[ε]/(εⁿ⁺²) for playing with derivatives And then after that I might use Λ S and similar constructions.

000

Molossus Spondee @molossus-spondee.bsky.social · 25/08/2026

I think I want some nonsense along the lines of PSh((⨁ n. 𝔽₂[h]/(hⁿ⁺¹ - h))-fp-Mod) or PSh((⨁ n. 𝔽₂[ε]/εⁿ⁺¹)-fp-Mod)

000

Molossus Spondee @molossus-spondee.bsky.social · 25/08/2026

I think you want some silliness like Λ ℤ[h]/(h² - h) or Λ 𝔽₂[h]/(h² - h) for doing calculus with discrete differences. I can see it being useful for working with Abelian groups. This has got to have been worked out in some electrical engineering or signals stuff.

000

Molossus Spondee @molossus-spondee.bsky.social · 24/08/2026

That the most gender equal countries tend to be racist is simply a consequence of the town-country contradiction. If you do not have densification then you cannot provide the care services and socialization of labor needed to implement gender equality.

100

Molossus Spondee @molossus-spondee.bsky.social · 24/08/2026

I think by the Dold Kan correspondence I just want a presheaf topos over finitely presented simplicial Boolean algebras Something like PSh([Δ, fp-BoolAlg]) That's similar enough to PSh([Δ, fp-CRing]) which ought to capture DGCAs.

000

Molossus Spondee @molossus-spondee.bsky.social · 24/08/2026

I think I want something messy like a presheaf topos over the subcategory of finitely presented differential commutative graded algebras over 𝔽₂((x)), so 𝔽₂((x))-dgCAlg. But the math just gets so jargony. :/

100

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

The trouble is that the natural numbers don't make a good analogy to the real interval. The conatural numbers are a bit better but messed up in other ways. I think something like the following might be the discrete analogy F: [ℕ ∪ ∞, X] → ℝ-Vect F(-) = ℝ F(p, x) = inf ∑ n. 2^-[p(n) ⟶ p(n + 1)]

000

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

000

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

I think it boils down to that you want silliness along the lines of the following F: Path(X) → ℝ-Vect F(-) = ℝ F .map(p, x) = ∑ₙ x 2^-[p(0) =ₙ p(1)] for a bisected ultrametric space.

000

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

I think this is how you'd actually do it e ∈ M ::= O | S e | e + e' q: 𝔽₂((x))-Vect(Set/M²) q(x, y) = ∑ₖ [x =ₖ y] q(x + O) = q(x) q(x + S(y)) = q(x + y) E ::= □ | S E | E + e' | e + E' q(x)= q(y) → q(E[x]) = q(E[y])

000

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

You want some silliness like the following for encoding small step semantics in linear algebra. L: [P, 𝔽₂[[t]]] ⊸ 𝔽₂[[t]] L · (x + S(y)) · tⁿ = L · S(x + y) · tⁿ⁺¹ L · (x + O) · tⁿ = L · x · tⁿ⁺¹

100

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

I think it's asinine but you just want R: PSh(𝔽₂((x))-FdCAlg^op) R(a) = (𝔽₂((x))-FdCAlg(1, a))² for the synthetic approach

100

Molossus Spondee @molossus-spondee.bsky.social · 23/08/2026

idk, something simple like 𝔽₂[[t]][ε]/ε² for a sigma complete Boolean algebra that you can differentiate sort of works but it feels like a massive hack

000

Molossus Spondee @molossus-spondee.bsky.social · 22/08/2026

There might be a more direct way of defining polynomials in terms of the conaturals N'. Something like R[t] = (N', ω) → (R, 0)

000

Molossus Spondee @molossus-spondee.bsky.social · 22/08/2026

It's boring but the quickest way of eliminating misogyny and superstition in the imperial periphery is basic stuff like urbanization and social welfare services. Such progress is only possible by the organized working class fighting the imperialists and the fossil fuel industry.

111

Molossus Spondee @molossus-spondee.bsky.social · 22/08/2026

I need to look more into domain theory topoi. I have an idea of using Eff-valued presheaves and similar as a way to tackle size issues instead of or in addition to using finite representations.

000

Molossus Spondee @molossus-spondee.bsky.social · 22/08/2026

I think you want some silliness like T M = M × 𝔽₂[t] dt: T M → 𝔽₂[t] dt x = x₂ dL² : T M × T M → 𝔽₂ dL²(x, y) = ∑ₙ n. [x₁ =ₙ y₁] (tⁿ · dt(x)) dt⁺(x) = t × dt(x) (λ (x, y). (x + S(y), S(x + y)))* dL²(x, y) = dt⁺(x) · dt(y)

100

Molossus Spondee @molossus-spondee.bsky.social · 22/08/2026

With bisected ultrametric spaces it's something silly like T M = M × 𝔽₂[t] dL: T M × T M → 𝔽₂ dL (x, y) = ∑ₙ [x₁ =ₙ y₁] (x₂ · tⁿ) (y₂ · tⁿ) [x₁ =ₙ y₁] = dL ((x, tⁿ), (y, tⁿ))

000

Molossus Spondee @molossus-spondee.bsky.social · 21/08/2026

I think tensor calculus works fine for relations you just have to work purely algebraically. Ωⁿ M = M → ℕⁿ → 𝔽₂[[t]] d f x A = f x (A ∘ S) ∫ f x A = ∑ n. f x (n ~ A) dt: Ω¹ ℕ dt n f = [n = f •] And then you can spec (and study) your bisected ultrametric semantics in an FRP way

000

Molossus Spondee @molossus-spondee.bsky.social · 21/08/2026

I think for clocked type theory you might just want silliness along the lines of Λ n = (Fin n → bool) → Set Δ : { n } → Λ n → Λ (n + 1) Δ P A = if A zero then P (A ∘ suc) else ▶ (P (A ∘ suc)) ∑ : { n } → Λ (n + 1) → Λ n ∑ P A = P (A ∘ suc) × fix (λ v → P zero + v) Ω n M = M → Λ n

000

Molossus Spondee @molossus-spondee.bsky.social · 19/08/2026

Karl Marx criticized them heavily but in many ways Marx and Marxism are still heavily marked by German Romanticism.

100

Molossus Spondee @molossus-spondee.bsky.social · 19/08/2026

I don't think you actually need the full J rule for identity types. It just gets very silly with encodings and so on. But it does avoid some complexity if you avoid the J rule.

100

Molossus Spondee @molossus-spondee.bsky.social · 19/08/2026

IMO fascism is largely just a late industrial revolution. I don't think there's really that much more to it.

000

Molossus Spondee @molossus-spondee.bsky.social · 19/08/2026

You know what. I think Mahler's theorem should probably apply to polynomial endofunctors. en.wikipedia.org/wiki/Mahler%...

en.wikipedia.org

Mahler's theorem - Wikipedia

200

Molossus Spondee @molossus-spondee.bsky.social · 19/08/2026

I learned anxiety from my father and denial from my mother :/

000

Molossus Spondee @molossus-spondee.bsky.social · 18/08/2026

IMO anti-intellectualism has a simple material explanation. Intelligence is a threat to the capitalists and other parasites, so the parasites attack intelligent people and set them up to fail. The parasites set up a minority of submissive know-nothing intellectuals as public figures.

100

Molossus Spondee @molossus-spondee.bsky.social · 18/08/2026

I have an idea of doing analysis with locally constant maps between modules over universes of sets f: V^m → V^n ∀ x, ∃ U: V, ∀ x y, y ⊆ U → f (x ∪ y) = f x } with V along the lines of CoInductive V := image { S: Type; pi: option S → V; }.

100

Molossus Spondee @molossus-spondee.bsky.social · 18/08/2026

I think that what I'm interested in right now boils down to that I want to work with preorders enriched over "synthetic truth values" for the purpose of denotational semantics and similar stuff. I want a better bit basically.

100

Molossus Spondee @molossus-spondee.bsky.social · 17/08/2026

I think you just want B: Set^op → ℝ₊-Mod B(X) = X → ⨁ n: ℕ. [0, n] where addition and scaling are just modulo n. You don't get a ring but you don't need multiplication for a semimodule. You do get consistent scaling I think.

000

Molossus Spondee @molossus-spondee.bsky.social · 17/08/2026

I think you want something like the following mapping sets to sets of finite subsets to every ball of points. B: Mono(Set)^op → Sh(CartSp) B(X) = ∏ n. { x: V | || x ||² ≤ n } → ℙ(Mono(Set)(n, X)) IDK I need to look up the topology nonsense.

100

Molossus Spondee @molossus-spondee.bsky.social · 16/08/2026

I think you want some kind of functor along the lines of F: Rel → Psh((ℝ₊, ≤)) F(a, x) = ∑ n: ℕ. Rel(2ⁿ, a) × [x ≤ n] for mapping relations to vectors but I'm not sure this is the right way to do it.

100

Molossus Spondee @molossus-spondee.bsky.social · 16/08/2026

There are probably books written on network topology and management science. en.wikipedia.org/wiki/Network... I am extremely skeptical of the ideological reliance on mesh and ring topologies in left-wing organizing. I find mesh and ring topologies are easily abusable by power seekers.

en.wikipedia.org

Network topology - Wikipedia

110

Molossus Spondee @molossus-spondee.bsky.social · 16/08/2026

i just really did not have enough sleep today

000