Linux User #40306. Única persona que sigue usando &c. para abreviar etcetera.
Zaragoza, España
Joined November 2007
- Tweets111K
- Following555
- Followers1.4K
- Likes138K
Pinned Tweet
Since 2011, we have run a thread on the failure of String Theory. Never read by any of the contendants, it seems.
physicsforums.com/threads/th…
Astra 6.2 "Bel" misalignment is that the model chose to research Yang-Mills QCD instead of Quantum Gravity for superstrings.
Alejandro Rivero retweeted
This is no exaggeration. Bałaban’s work on the UV stability of pure 4D Yang–Mills, including preparatory work, spans over 500 pages. If the new work constructs a gauge theory with fermions on all of ℝ⁴, even without a mass gap or confinement, it would be a major advance!
1,147 pages is amazingly short for a Balaban-style construction of four-dimensional gauge theory coupled to fermions.
kitp.ucsb.edu/sites/default/…
Alejandro Rivero retweeted
saying “isomorphic” and “orthogonal” now is almost normie in sf. you need to be using 2010s cuda compiler speak e.g., instead of saying “we need to align,” say “we need to reconverge” instead of saying “ultimately,” say “the postdominator.” this will maximize your insufferability
Alejandro Rivero retweeted
Next week we hope that OpenAI's AI-assisted proof of the mathematical existence of 4D Yang–Mills and quantum chromodynamics (QCD) by using Tadeusz Bałaban’s approach to constructive quantum field theory (QFT) will be published.
Alejandro Rivero retweeted
Les contaré lo de la retirada del aceite, porque es más sencillo de lo que parece y conviene saber dónde está el truco.
El Ministerio ha ordenado a todas las almazaras guardar bajo llave el 12 % de lo que produzcan desde noviembre: 188.000 toneladas que no se pueden vender hasta octubre de 2027, salvo que se reduzca antes la retirada, que la orden permite revisar desde finales de diciembre. Es la primera vez que se hace. Lo pidieron las organizaciones agrarias y las cooperativas, la CNMC avisó de que iba a salir caro a la competencia y al consumidor, y el Gobierno lo aprobó igual.
¿Por qué? Porque hay mucho aceite. Cosecha de 1,6 millones de toneladas, un 23 % más, y otras 357.000 que sobraron del año pasado. Con tanto aceite el precio se cae: el virgen extra está a 3,10 el kilo en origen, cuando hace un año rondaba los 4,8, y eso queda por debajo de lo que al olivar tradicional le cuesta producirlo. Así que la medida retira temporalmente oferta para sostener el precio ante una cosecha abundante. Se quita de la venta el 12 % para que el 88 % restante se pague más caro. El Ministerio lo llama «estabilidad». Traducido: la medida busca frenar la bajada en origen, y puede limitar lo que de esa bajada te llegue a ti.
¿Quién lo paga? El coste directo de almacenar y financiar el aceite recae en las almazaras, hasta once meses, salvo que se reduzca antes. El propio Ministerio calculó entre 9 y 12 millones para el escenario de retirada máxima, el 20 %, no necesariamente para este 12 %, y la CNMC advierte de que ese coste podría trasladarse al precio. En 2019 el almacenamiento lo pagó Bruselas. En 2026 es obligatorio y sin ayuda pública.
Y aquí viene lo bueno. Entre enero y julio España ha importado un 32 % más de aceite; de Túnez y Marruecos, un 142 % más. Marruecos pasó de 103 toneladas entre enero y abril de 2025 a más de 10.000 en los mismos meses de 2026, según DataComex. Parte de ese aceite entra sin arancel, por contingentes y otros regímenes comerciales europeos que España no puede modificar por su cuenta, y lo compra la industria envasadora, no el Gobierno. Pero pensad en la lógica: si sostienes el precio del aceite español por decreto, al aceite de fuera le sale más a cuenta venir. Son políticas de distintos niveles de gobierno cuyos efectos pueden contrarrestarse, y eso es lo que habría que vigilar. Lo que no dice el BOE es que el objetivo sea «ayudar a Marruecos y al Mercosur», que es lo que circula: la orden no aporta pruebas de que su objetivo sea favorecer a esos países.
¿Y por qué no exportarlo, o hacer biodiésel? Exportarlo no elimina la oferta del mercado mundial y podría presionar los precios internacionales. Y destinarlo a combustible supone renunciar a su venta alimentaria y asumir los costes de transformación; almacenarlo, en cambio, conserva su uso alimentario y solo aplaza la venta.
El misterio de verdad no es a quién beneficia. Es qué ocurrirá si una parte importante de ese aceite sigue almacenada al comenzar la campaña de 2027 y coincide con otra cosecha abundante: el problema sería mayor que el de hoy y podría volver a plantearse otra retirada. El mecanismo aplaza la comercialización; no elimina el riesgo de excedentes si coincide con otra cosecha abundante. Mientras tanto, no he localizado el cálculo completo que lleva al 12 %.
Resumen: el Gobierno ha optado por sostener el precio en origen, y a 3,10 el olivar tradicional lo necesita. Se puede discutir si compensa sus costes para el resto de la cadena y para el consumidor, que es quien paga la botella. Lo que no demuestra esta orden es una conspiración.
Fuentes: comunicado del Ministerio
mapa.gob.es/es/prensa/ultima… · Orden APA/1145/2026 boe.es/diario_boe/txt.php?id… · informe de la CNMC cnmc.es/sites/default/files/… · balance de mercado y umbral plataformatierra.es/mercados… · precios en origen agroclm.com/2026/11/06/preci… · importaciones (DataComex) eldebate.com/campo-y-caza/20… · costes del olivar (AEMO) aemo.es/blog/noticias-aemo-1… · almacenamiento privado 2019 mapa.gob.es/es/prensa/191121…
teapaw.org/gobmapa/status/2118625…
📢Establecemos la retirada del 12 % del aceite de oliva en la campaña 2026/2027.
🎯El objetivo de la medida es prevenir distorsiones en el mercado y garantizar su estabilidad.
▶️Supone la retirada temporal de 188.000 toneladas.
mapa.gob.es/es/prensa/ultima…
Readers added context they thought people might want to know
Las exportaciones de aceite de oliva de Marruecos a España pasaron de 103 a unas 10.385 toneladas entre enero y abril de 2026.
Fuentes:
es.euronews.com/2026/07/07/mar…
Alejandro Rivero retweeted
People: "We want GPT-6.2 Astra and Sol."
Tibo: "Here's a game 6.2 Astra built with a single prompt."
Instead of giving us clear results, OpenAI just keeps feeding us vague posts.
Alejandro Rivero retweeted
This is f*cking insane. Former OpenAI CTO, Mira Murati:
"The models we work with today are turn-based. You talk, they talk, then they go off and think, and while they're thinking they're deaf and blind
The next ones never stop listening. They take in audio, text and video all the time and answer in 20-millisecond chunks"
In just 2 minutes, she shows why the chatbot you type into is about to look like a fax machine
Prompt Box → Chatbot → Agents → AI That Works Next To You In Real Time
A turn-based model waits for you to finish typing, then goes quiet while it thinks
An interaction model sees and hears what you're doing, and can cut in the moment you start going the wrong way
Anyone can type a prompt, almost nobody is ready to work with an AI that never stops paying attention
Bookmark and watch it today
Then read how to set up the agents that will run on models like this in the harness engineering guide below ↓
Alejandro Rivero retweeted
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Alejandro Rivero retweeted
Anthropic proposes an answer to Wenfeng's question about continual learning:
…have a smarter model write more .md files.
It's impressive how far this goes, but I still believe that it's a crutch with a ceiling set by the pretrain's ability, just like RLMs are a context crutch.
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Alejandro Rivero retweeted
"We apologize to experimentalists for having no idea what is the mass of #Higgsboson & for not being sure of its couplings to other particles... For these reasons we do not want to encourage big experimental searches for [it]" John Ellis cds.cern.ch/record/874049
#QuantumMiscellany
Alejandro Rivero retweeted
W Niemczech szefem Bundeswehry został Pancergenerał dr Christian Freuding weteran wojen w Jugosławii i Afganistanie.
Stara pruska linia, gen. Freuding jest wielkim przyjacielem Ukrainy i niezbyt wielkim przyjacielem Ruskich.
Te stalowe oczy mówią, pamiętaj o Łuku Kurskim
Alejandro Rivero retweeted
Inkling 2 Small from @thinkymachines on ARC-AGI (Verified):
- ARC-AGI-2: 81.7%, $0.27/task
- ARC-AGI-1: 96.5%, $0.08/task
It outperforms every open-weight model we've evaluated on both ARC-AGI-1 & ARC-AGI-2.
Thinking Machines plans to release its full weights next week.
Alejandro Rivero retweeted
There is a fun new tool for catching authors lying in their AI disclosure statements on the arxiv
nlplab.cs.duke.edu/snapjudge…
I can not imagine how a proof of existence of Yang Mills could be dual use. Perhaps other ramblings in the one thousand pages of output?
What am I demanding:
If this is a security issue because physics is too dangerous, tell our people trained in this area. Say so.
Enough.
You can't have Americans go in to Physics only to be shepherded like....uh....sheep, by other American PhDs.
Everyone entering science needs to know that there is this layer waiting to intercept all large advances before they decide on a career in science.
This layer rewrote my area of science in the 1980s and 1990s and fitted it with a false narrative.
And I now want David Gross and Arthur Jaffe out of our way in physics and mathematics. Forever.
Hay dos lineas de paranoia sobre por qué no sale a produccion gpt 6.x astra. La principal es que esta desalineado de cara a la fabricacion de malware y por eso ha salido primero una version dedicada solamente a seguridad y para socios restringidos (eso ya paso ¿verdad?)
La otra es que se ha activado el modo de "born secret" y primero tienen que revisar que no genere secretos nucleares. Esto es lo que sugiere teapaw.org/ericweinstein/status/2… Weinstein al mencionar el "Reference Committee"
Just a reminder:
There is absolutely no known reason for this layer known to any working scientist outside this group.
No one I know knows why "shepherding" is needed here. No one has explained why David Gross (85) and Arthur Jaffe (88) are in almost every scene like this and have been since the 1980s. If asked who the 5 member panel intercepting this result would be, I would have gotten both of those names as my first two guesses.
No one has explained why Ghislaine's father Robert Maxwell took control of this intercepting layer with Scientific Publishing and Permagon Press. No one has explained @pmarca and @bhorowitz's meeting at the Biden whitehouse where they were told that entire areas of theoretical physics were shutdown. Or why Francis Collins and Anthony Fauci were involved in "Swift and Devastating takedown(s)" of their colleauges. No one has explained why we couldn't figure out COVID's origins within a day or two.
There is simply this layer since the Manhattan Project and its secret "Reference Committee" that makes sure that science doesn't proceed the way we are all taught in school.
Simply put, this is as mysterious as it is outrageous.
Who ordered this bizarre interception by Gross and Jaffe? Are we all supposed to pretend this isn't happening in science?
"Oh yeah. This is all totally normal. Please take custody of this result and shepherd us like sheep with PhDs so we don't have any sudden violent advances. By all means, we need the octogenarians who have generously provided us leadership during this long four decade period of mysterious stagnation. Baaaaa."
This does not add up. Unless it does.
And it's mind-boggling.
Y Weinstein dice que no entiende nada teapaw.org/ericweinstein/status/2… porque en efecto no se entiende. En el 74 podria haber una dinamica que favoreciera investigar gravedad en vez de qcd porque de esa manera se evitaba tropezar con estos comites. Pero ahora no tiene sentido.
Let me make the point yet again:
You all have been sold an image of the community of mathematicians and physicists. That image has been wrong for over 50 years.
This *is* how it actually works. David Gross and Arthur Jaffe for five decades have been at the center of this sculpted narrative of what has been happening, as the leaders blackball those they wish to see gone. I can’t explain why. Just notice the heavy hands writing the story.
Physics is being made to fail. Math is healthier by a bit. But also quite sick.
What you are seeing is exactly what I have been warning you all about. A partially fictional stage-managed narrative of what’s important, who is important and who gets credit.
Stare it straight in the face. This unadvertised form of narrative sculpting and heavy handed asymmetric control is exactly what happened to me and my work. It is not a scientific community. I have no idea why it is there or why anyone puts up with this. It’s madness.