DarkImmanuel 🎃 retweeted
CHICAS ME HIZO LOVEBOMBING Y DESPUÉS ME GHOSTEÓ Y AHORA ME HACE ORBITING QUE HAGO? LE HAGO BREADCRUMBING O BENCHING?
DarkImmanuel 🎃 retweeted
mi ex: superada
esa mujer con quien tuve una conexión increíble pero no fuimos nada: tema sensible
DarkImmanuel 🎃 retweeted
nate silver, american statistician born in east lansing in 1978, built a forecasting model in his apartment in 2008 that correctly predicted the presidential election in 49 out of 50 states. every major TV network had a room full of analysts. he had a laptop and a probability distribution.
ESPN acquired FiveThirtyEight for $30M+. then shut it down. he rebuilt alone as Silver Bulletin.
his entire career is one argument: prediction is not a point. it is a distribution. in 2026, every LLM running on $4.6T worth of infrastructure is proving him right one token at a time.
this october 2026 paper - independent research, not affiliated with any institution - shows the mathematics that every frontier model is running underneath the surface that almost nobody understands.
p(x_{t+1:T} | x_{1:t}) = ∏ p(x_k | x_{<k}). LLMs don't predict one future. they generate a probability distribution over millions of possible token sequences simultaneously. every response you read is one sample drawn from that distribution.
p(x_{t+1} | x_{1:t}) = softmax(z_{t+1}/τ). τ is temperature. when τ → 0: the model collapses to one deterministic answer. when τ = 1: full probability distribution fires. the "creativity" setting in every AI product is this one variable.
uncertainty compounds with each step forward. H(X_{t+k} | X_{1:t}) ≈ H(X_t) + k·σ². entropy grows linearly with prediction horizon. at 10 tokens ahead: manageable. at 100 tokens: exponential spread. at 1,000 tokens: the model is navigating a fog it cannot see through.
attention weights at horizon k=80: drop to 10^-3. the model barely perceives 80 tokens ahead. it is not predicting the future. it is pattern-matching the immediate past and calling it intelligence.
future states don't scatter randomly. they cluster into coherent semantic manifolds in latent space. scientific continuations form one attractor. economic narratives form another. creative trajectories form a third.
the model holds all of them simultaneously until forced to collapse into one output.
silver said in 2012: "the problem is not that we have too little data. it's that we don't understand probability."
the AI industry raised $4.6T building GPT-4, claude, gemini, and grok. this paper shows those same systems are running full probabilistic machinery underneath that almost nobody is using correctly.
silver was right in 2012. the math was there since kolmogorov in 1933. the compute to run it arrived in 2024.
save this.
hoy bailaremos con @dabeull y veremos a la siempre sexy HolyBrune
2
124
Sólo quiero que mi spookysad girl disfrute está canción tanto como yo.
Fontaines DC los amo y el siguiente año los veré
118
DarkImmanuel 🎃 retweeted
BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2026 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel to Robert E. Hall and Michael Woodford “for microeconomic foundations of macroeconomic stabilisation.”
#NobelPrize
DarkImmanuel 🎃 retweeted
BREAKING NEWS
The 2026 #NobelPrize in Literature is awarded to the Canadian author Anne Carson, “for her bold and inventive oeuvre that, in playful dialogue with the classical tradition, has created new forms for contemporary literature.”
DarkImmanuel 🎃 retweeted
BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2026 #NobelPrize in Physics to Francis Halzen “for decisive contributions to the IceCube Neutrino Observatory and the discovery of high-energy neutrinos of astrophysical origin.”
DarkImmanuel 🎃 retweeted
On Mezzanine, Massive Attack tried to escape trip-hop. They nearly tore themselves apart and made its defining document instead pitchfork.com/reviews/albums…
I love games, and I love music just as much.
Some albums shaped who I am, so I made my9albums.
Pick the 9 albums that made you, turn them into a poster, and share it.
my9albums.com
1
65
DarkImmanuel 🎃 retweeted
Hay una caída de la fecundidad en México, la CDMX está casi a los niveles de Seúl, las mujeres están decidiendo no embarazarse; y hay voces culpando al aborto legal de este fenómeno como si las mujeres estuvieran obligadas a ser madres. Pero omiten el impacto de la inseguridad, la precariedad laboral, la ausencia de un sistema de cuidados, la inestabilidad económica, la violencia de género, el aumento del abuso sexual infantil, la brecha salarial y la feminización de los cuidados como factores que inciden en esta decisión.
Estamos viendo el reflejo de la autonomía reproductiva, las mujeres no le deben hijos a nadie ni son máquinas reproductoras de fuerza laboral.
el nuevo disco de Jungle es como una sola canción de una hora
Our new album SUNSHINE is available on iTunes for just £5.99, for a limited time only 🌞😎
jungle.ffm.to/itunes
42

























