Roger Senserrich
LightNews LightNews
banner
4freedoms.es
Roger Senserrich
@4freedoms.es
5.5K followers 810 following 5K posts
Poldavo militante, pro-naumaquias. Amante de la burocracia. Substack sobre política americana en https://4freedoms.substack.com/.
Posts Replies Media Videos
Reposted by Roger Senserrich
4freedoms.es
En Four Freedoms, cuando la administración Trump decide que leer la ley en voz alta es una conducta sediciosa y mereces ser ejecutado por ello. https://www.4freedoms.es/p/exigiendo-lealtad?r=4b3tr&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false
November 26, 2025 at 6:01 PM
Reposted by Roger Senserrich
jamellebouie.net
precisely my thought. the president and his dumbass republican allies put these men and women into harm’s way for a political stunt and look what has happened
aubreygilleran.bsky.social Aubrey Gilleran @aubreygilleran.bsky.social · 15h
Whether this was random street crime or targeted political violence, I think it was bad for the president to unnecessarily put the National Guard in harm’s way like this. They simply don’t need to be there.
November 26, 2025 at 8:53 PM
Reposted by Roger Senserrich
wsj.com
Breaking: The two National Guard members shot near the White House have died. A suspect is in custody and is "severely wounded," President Trump said.
Two National Guard Members Killed in D.C.
A suspect was also shot and is in custody, U.S. officials say
on.wsj.com
November 26, 2025 at 9:04 PM
4freedoms.es
En Four Freedoms, cuando la administración Trump decide que leer la ley en voz alta es una conducta sediciosa y mereces ser ejecutado por ello. https://www.4freedoms.es/p/exigiendo-lealtad?r=4b3tr&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false
November 26, 2025 at 6:01 PM
Reposted by Roger Senserrich
giraldeo.es
Oye mirad qué chulo esto nuevo de @renfeoperadora.bsky.social Todos los trenes de cercanías de España mapeados, con información de marcha a tiempo real y datos en abierto. ¡Y se mueven! 🚈
tiempo-real.renfe.com
November 26, 2025 at 5:17 PM
4freedoms.es
No me creo conspiraciones, PERO…
saeedwkhan.bsky.social Saeed Khan @saeedwkhan.bsky.social · 21h
Just a reminder:

The night Epstein "committed suicide".

-He was on suicide watch.
-The guards "fell asleep" and failed to complete more than 75 mandatory checks on him.
-The cameras monitoring him "stopped working"
-Prosecuting the guards was "deferred"

www.npr.org/2023/06/27/1...
Negligence, staff failures led to Jeffrey Epstein's suicide, a DOJ report says
The disgraced financier who was jailed for sex trafficking charges, killed himself at New York's Metropolitan Correctional Center in August 2019. He died before standing trial.
www.npr.org
November 26, 2025 at 5:50 PM
Reposted by Roger Senserrich
samwang.bsky.social
West Point plaque: "Our code of military obedience requires that, should orders and the law ever conflict, our officers must obey the law."
November 25, 2025 at 10:40 PM
4freedoms.es
You'll be visited by three spirits.

The three spirits:
November 26, 2025 at 2:14 AM
4freedoms.es
Atlanta? Who the hell wants to move to a city that has an aquarium and the Jimmy Carter library as main cultural amenities? The city is utterly charmless.
November 26, 2025 at 2:07 AM
Reposted by Roger Senserrich
gbrockell.bsky.social
Ms. Kuyateh was one of 19 third-country nationals sent to Ghana on Nov. 6. At least 61 TCNs have been removed there since September.

Others report similar treatment. Nigerians were forced to cross into Togo w/o documents, rendering them essentially stateless

www.nbcwashington.com/investigatio...
Woman deported from Maryland shown on video being dragged in Ghana
Rabbiatu Kuyateh was detained this summer at her annual ICE check-in, her son said. She moved to the D.C. area 30 years ago as she fled civil war in Sierra Leone.
www.nbcwashington.com
November 26, 2025 at 1:04 AM
Reposted by Roger Senserrich
nytimes.com
The six Democrats in Congress who recorded a video informing troops that they could refuse illegal orders said they were being investigated by the FBI. The inquiry is part of a campaign by the Trump administration to seek retribution against the president’s perceived political enemies.
Democrats Say F.B.I. Is Investigating Them Over Illegal Orders Video
The move came one day after the Pentagon said it was investigating Senator Mark Kelly for his participation in the video.
nyti.ms
November 25, 2025 at 9:10 PM
4freedoms.es
The dirty secret that urbanist Bluesky doesn't want to tell you is that urban planning is a solved issue. We have all the answers. We found the one true solution of the perfect urban form.

This is it. Out job now is making the world more similar to this.
November 25, 2025 at 6:53 PM
Reposted by Roger Senserrich
razzball.bsky.social
A healthcare system where we hope someone wants our autograph
November 25, 2025 at 5:34 PM
4freedoms.es
Señoros de la FORMULA UNO:

Haced la carrera substituta en Montjuic.

Venga, no hay huevos.
eljaviparla.bsky.social Don Javier de Parla @eljaviparla.bsky.social · 1d
Otro gran éxito del PP de Madrid que va a costarnos un puto riñón a los madrileños.
November 25, 2025 at 6:22 PM
4freedoms.es
For all the bazillions of dollars Microsoft is investing in AI - could they spare just a tinsy bit of cash to make Word not suck so much when formatting documents?

"Oh, now I am adding spaces here! The default font when pasting is Times New Roman size 8 now! I am bolding everything !"
November 25, 2025 at 4:27 PM
4freedoms.es
You’ve been kidnapped. The characters from the last TV show you watched are trying to rescue you. Who’s coming to save you?

I'll be home for dinner, and my kidnappers are going to die horrible, painful deaths.
November 25, 2025 at 2:58 PM
4freedoms.es
Alexander Isak me parecía un delantero decente para la Real Sociedad, pero no un jugador que valiera la millonada que pagaron por llevárselo.

El Liverpool pagando 125 minolles de libras por alguien que sería suplente en el Atlético es muy divertido. www.nytimes.com/athletic/683...
There are 125 million reasons why Alexander Isak is becoming a big problem for Liverpool
www.nytimes.com
November 25, 2025 at 2:41 PM
4freedoms.es
Mirando al azar precios de RAM, compré 32 GB de DDR4 hace un año por $54. La misma memoria está a $154 hoy.

Quería ampliar el PC nuevo (DDR5) de 32 a 64 GB, y me sale más a cuenta comprar una PS5 Pro.
November 25, 2025 at 1:46 PM
Reposted by Roger Senserrich
joshtpm.bsky.social
At this point X is like 1950s Vienna, a whole place made up entirely of spies, running ops against other spies, with no non-spy people. Or maybe like a 1971 Black Panthers convention where 3/4 of the attendees are conintelrpo.
November 24, 2025 at 3:50 AM
Reposted by Roger Senserrich
danglinghemmie.bsky.social
it's happening dot gif
November 24, 2025 at 7:46 PM
Reposted by Roger Senserrich
nachonachete.bsky.social
"MTG y su extraña historia de quijotismo reaccionario".

👇👇
4freedoms.es Roger Senserrich @4freedoms.es · 2d
En el Four Freedoms de hoy, la peculiar carrera política de Marjorie Taylor Greene, alguien que ha ido de ser la republicana más extremista del congreso a dimitir tras un súbito ataque de lucidez... sin cambiar una sus opiniones o ideas.

www.4freedoms.es/p/la-honesti...
La honestidad de la locura
Marjorie Taylor Greene anuncia su dimisión
www.4freedoms.es
November 24, 2025 at 10:03 PM
4freedoms.es
The right order is:

1. Empire Strikes Back
2. Rogue One.
3. A New Hope.
4. Revenge of the Sith.
5. Last Jedi.
6. Return of the Jedi.
7. Solo.
8. Phantom Menace.
9. Force Awakens.
10. Attack of the Clones.

I refuse to accept the putrid turd that is ep. IX as canon. It is unwatchable.
danlehner.bsky.social paul khruangbin (Dan Lehner) @danlehner.bsky.social · 2d
haven’t seen ROS or Solo but this seems like a pretty ball-knowing list from Pritzker, particularly if he volunteered the name “Star Wars” rather than “A New Hope”
So, we asked Pritzker to weigh in on the go-to debate for every "Star Wars" nerd: Which movies are the best? Here's his ranking of the 11 major films in the series: 1. "The Empire Strikes Back" (1980) 2. "Star Wars" (1977) 3. "Return of the Jedi" (1983) 4. "Rogue One" (2016) 5. "The Last Jedi" (2017) 6. "Revenge of the Sith" (2005) 7. "The Force Awakens" (2015) 8. "Solo" (2018) 9. "The Phantom Menace" (1999) 10. "Attack of the Clones" (2002) 11. "The Rise of Skywalker" (2019)
November 24, 2025 at 7:45 PM
Reposted by Roger Senserrich
ketanjoshi.co
Deeply absurd. This Google PDF published on a blog (arxiv, not peer reviewed) claims an LLM is "PhD level" but in most cases the MAJORITY of reference URLs were invalid or inaccessible.

A PhD sitting down and just fabricating >50% of sources = career ending

arxiv.org/abs/2511.11597
CLINB: A Climate Intelligence Benchmark for Foundational Models Michelle Chen Huebscher1 , Katharine Mach2 , Aleksandar Stanić1 , Markus Leippold1,3, Ben Gaiarin1 , Zeke Hausfather4 , Elisa Rawat , Erich Fischer5 , Massimiliano Ciaramita1 , Joeri Rogelj6 , Christian Buck1 , Lierni Sestorain Saralegui1 and Reto Knutti5 1Google DeepMind, 2University of Miami, 3University of Zurich, 4Stripe, 5ETH Zurich, 6 Imperial College London Evaluating how Large Language Models (LLMs) handle complex, specialized knowledge remains a critical challenge. We address this through the lens of climate change by introducing CLINB, a benchmark that assesses models on open-ended, grounded, multimodal question answering tasks with clear requirements for knowledge quality and evidential support. CLINB relies on a dataset of real users’ questions and evaluation rubrics curated by leading climate scientists. We implement and validate a model-based evaluation process and evaluate several frontier models. Our findings reveal a critical dichotomy. Frontier models demonstrate remarkable knowledge synthesis capabilities, often exhibiting PhD-level understanding and presentation quality. They outperform “hybrid" answers curated by domain experts assisted by weaker models. However, this performance is countered by failures in grounding. The quality of evidence varies, with substantial hallucination rates for references and images. We argue that bridging this gap between knowledge synthesis and verifiable attribution is essential for the deployment of AI in scientific workflows and that reliable, interpretable benchmarks like CLINB are needed to progress towards building trustworthy AI systems. Total Reference URLs Generated claude-opus-4-1 claude-sonnet-4 gpt-5 hybrid gemini-2.5-pro gemini-2.5-flash o3 0.0 0.2 0.4 0.6 0.8 1.0 Proportion Reference URL Status hybrid gemini-2.5-pro claude-opus-4-1 o3 gemini-2.5-flash claude-sonnet-4 0 200 400 600 800 1000 Count of URLs Total Image URLs Generated hybrid claude-opus-4-1 gemini-2.5-flash gemini-2.5-pro claude-sonnet-4 o3 0.0 0.2 0.4 0.6 0.8 1.0 Proportion Image URL Status Status OK INACCESSIBLE_CONTENT INVALID_URL ERROR Figure 3 | Number of reference (top), and image (bottom), URLs and their status. Ablations We perform several ablation studies with the autorater (Table 4). Notably, removing the question-specific rubrics from the prompt changes the results only in the bottom half, with the Hybrid answers overtaken by Gemini 2.5 Flash and Claude Sonnet 4. This suggests that the additional resolution provided by the rubrics applies primarily to the kind of responses used to develop the rubrics. Or, in other words, that rubrics are far from complete. Hence, it is important that rubrics adapt to new data as better models become availab A New Expert-Grounded Benchmark for Scientific AI We introduce CLINB, a benchmark for modelbased evaluation of frontier models on complex, multimodal scientific communication. Its core is a new dataset of real-world climate questions paired with data-driven, question-specific evaluation rubrics, curated and validated by leading climate scientists through a novel three-phase, human-in-the-loop process.2 PhD-Level Synthesis vs. Attribution Failures Frontier models demonstrate remarkable knowledge synthesis, often exhibiting a PhD-level understanding. However, this performance masks a critical inadequacy in grounding. We report substantial hallucination rates for references (10% to 25%) and even more failures for images (50% to 80% in certain settings), exposing a major gap between synthesis and verifiable attribution. Insights into Human-AI Collaboration Dynamics Autonomous frontier models surpass ’hybrid’ answers (curated by experts using weaker AI assistance), revealing the assisting model’s capability—not human oversight—as the primary bottleneck. Counter-intuitively, highly motivated non-specialists (our ’Advocates’) who deeply engage with AI tools can produce higher-quality answers than domain experts who engage less with AI during answer curation. A Validated Methodology for Scalable Oversight We validate a rigorous, rubric-based autorater. Ablation studies demonstrate that structured prompts and automated evidence-checking are essential for mitigating inherent LLM judge biases. This process is hampered by inaccessible sources (up to 50%). Furthermore, we identify evaluation challenges, including model familiarity bias in human raters and the limitations of rubrics to generalize across models.
November 24, 2025 at 7:36 PM
4freedoms.es
This is my most reactionary opinion, but I hate, hate, hate how people dress down when flying.

I am not that old, but I remember in the 80s my parents mentioning you had to look nice when taking a plane. To this day, I avoid wearing sneakers or T-shirts when flying (decent shoes or polos instead).
atrupar.com Aaron Rupar @atrupar.com · 2d
Sean Duffy: "Dressing with respect -- whether it's a pair of jeans and a decent shirt, I would encourage people to maybe dress a little better which encourages us to maybe behave a little better. Let's try not to wear slippers and pajamas as we come to the airport."
November 24, 2025 at 7:39 PM
Reposted by Roger Senserrich
kairyssdal.bsky.social
The number of ways this is disgraceful behavior from the Secretary of Defense is...a lot.

Service members do have a duty to refuse illegal orders

All Americans have a 1st Amendment right to say that out loud.
ronfilipkowski.bsky.social Ron Filipkowski @ronfilipkowski.bsky.social · 2d
Hegseth is now threatening to recall Sen. Mark Kelly to active duty in order to court martial him.
November 24, 2025 at 5:29 PM