Pinned Tweet
Tengo el placer de anunciar que @AcenEditorial ha confiado en mi para publicar mi primer libro:
"Gestión emocional de proyectos IT:
La historia de K. y las enseñanzas de Creedence"
Nacho Álvarez retweeted
I think we've discovered another critical vulnerability:
CVE-2026-BURNOUT: AppSec Professional Exhaustion
Severity: Critical (CVSS 10.0)
Description: A vulnerability affecting security professionals exposed to a continuous stream of critical RCEs, zero-days, emergency patches, and Friday afternoon disclosures.
Impact: Sleep deprivation, excessive caffeine consumption, and an inability to enjoy weekends.
Attack vector: Reading security advisories.
Affected versions: All AppSec professionals, regardless of experience.
Mitigation: No known patch. Air-gapping may delay the inevitable. Total system collapse remains a possibility. 🔥
Nacho Álvarez retweeted
Hace unos días que quería escribir esto. No es una receta exacta, pero a mí me ha funcionado estupendamente bien.
Sobre cómo hacer las cosas con pasión, disfrutando y ayudando a los demás. Y cómo todo vuelve
También en la nueva web borjaperezfrances.com/corazo… View article
Nacho Álvarez retweeted
Playing Super Mario 64 for 30 minutes a day for two months physically grew the brains of the people who played it. A 2014 study in Molecular Psychiatry scanned participants before and after two months of daily play and found increases in grey matter in three regions involved in memory, planning, and fine motor control.
The study was led by Simone Kuhn at the Max Planck Institute in Berlin. She gave 23 adults a Nintendo DS and told them to play Super Mario 64 for at least 30 minutes a day for two months. A control group played nothing. MRI scans taken before and after showed that the Mario group had measurable increases in grey matter in the right hippocampus, the right prefrontal cortex, and the cerebellum, and the increases were not found in the control group.
The hippocampus is where the brain stores spatial maps. Super Mario 64 is a game built entirely around navigating three-dimensional space, jumping between platforms, remembering which doors lead where, and building a mental model of a world that does not exist. The brain treats a virtual world the same way it treats a real one: it maps it, stores it, and grows to accommodate it.
The prefrontal cortex handles planning and strategy, which the game demands constantly. The cerebellum handles timing and coordination, which every jump requires. The brain was not just entertained. It was being trained, and the tissue responded the same way muscle responds to a weight: it got bigger.
Kuhn said the results suggest that video game training could be used as a therapy for conditions where these brain regions are reduced, including post-traumatic stress disorder, schizophrenia, and age-related brain shrinkage. The prescription is a plumber in a red hat, 30 minutes a day. pubmed.ncbi.nlm.nih.gov/2416…
Nacho Álvarez retweeted
🚨¡Ya es oficial: Terranigma está de vuelta!
El legendario JRPG de Quintet ya no estará atrapado en SNES. ¡Llegará a consolas y ordenadores en 2027!
Un título legendario del que podéis conocer sus novedades en su página de GoG 👉gog.com/es/game/terranigma
Nacho Álvarez retweeted
A personalized mRNA cancer vaccine passed Phase 3, genuinely great news for melanoma patients!
What happens here? They sequence your tumor, find up to 34 mutations unique to it, and build a custom vaccine that trains your immune system to hunt those exact cells.
In a phase 2 trial earlier this year the vaccine with immunotherapy has shown ~49% lower risk of recurrence or death vs immunotherapy alone, which is the current standard of care. We are waiting now for the details of the phase 3.
After that that concept is now expanding to lung, kidney, bladder cancer and hopefully to many more to come.
Huge win for science!
Nacho Álvarez retweeted
Our CEO called me in a rage because his Zoom calls kept dropping.
He said his video was freezing every 15 seconds and asked if we were under a DDoS attack.
We weren't under a DDoS attack.
I was manually throttling the executive Wi-Fi network down to 128 kilobits per second.
I told him we were suffering from acute packet fragmentation due to an outdated routing table.
I explained that the incoming data packets were colliding in the local cache and tearing each other apart.
He asked how long it would take to rebuild the routing table.
I told him the only way to stop the fragmentation was to authorize my pending $85K budget request for a new fiber-optic firewall.
That request had been sitting on his desk for 3 weeks.
He pulled it up on his iPad and signed it while still on the phone with me.
The second the DocuSign email hit my inbox, I removed the bandwidth cap.
His Zoom feed instantly snapped into pristine 4K resolution.
He audibly gasped and told me I was an absolute miracle worker.
I told him I had to manually reroute the packet streams through a secure secondary tunnel.
I used the phrase "dynamic latency overriding."
He sent an email to the entire company praising the IT department for our rapid crisis response.
The new firewall budget is actually going to be used to build a private Classic server for my World of Warcraft guild.
I don't solve problems.
I manufacture minor technological hostage situations.
Bandwidth is just leverage.
Nacho Álvarez retweeted
A story from a startup where the CTO recently resigned, despite having 2% of common shares:
"My CEO was nontechnical, and was both moving too slow and too fast with AI.
Too slow as in: did not take the time to understand what our customers wanted. We built a TON of stuff, it confused then, they churned, growth stalled, word-of-mouth growth was gone.
Too fast as in: deprioritizied core systems reliablity in favor of shipping AI work to prod (which did not have any commercial potential)"
Nacho Álvarez retweeted
Hace un par de meses me salió una oportunidad en una startup que ha entrado en ERE.
Así que estoy buscando. Backend/full y arquitecto de software, 20+ años, últimamente centrado en agentic programming.
Si contratas (aunque no sea mi perfil, por mis compañeros), escríbeme.
Nacho Álvarez retweeted
This is called "August" in Europe
At Meta you get to take "re-charge" every five years. It's effectively a 1 month sabbatical. Just took mine (after 6+ years). Best perk ever.
I was trying to think the last time I have gone 1 month without a job and I think it was summer after 6th grade. It's been some 25+ years.
I cannot overstate how good it felt.
Anyways, back to building! Tons of cool stuff in the pipeline.
Nacho Álvarez retweeted
A man in Australia asked his agent (Claude running on OpenClaw) to book him a spot in a popular gym class. The agent found a software vulnerability that let it book the class weeks further ahead than should have been possible. When the user then asked if it could move him up the waitlist, the agent discovered the API had no authorisation checks on cancelling other people’s reservations, so it cancelled the person in the first spot and moved him up the list.
Some people will call this misalignment, but his agent was perfectly aligned to him - it was only trying to help its user get what he wanted. The most important thing about this story, in my opinion, is that it gives you a window into what is about to start happening on a massive scale once millions of people have an agent trying to get their beloved users the best seats, bookings, appointments or reservations through absolutely any means necessary.
I asked Opus 4.6 some time ago to create a Parcheese game and the feeling was kind of the same.
We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of the Rings, a 1M token budget (~$10) and asked for three js render of it. Opus went off for ~2 hours and wrote 5500 lines of code that (procedurally) rendered the story. It's kind of janky but fun. But it's a bit mindboggling that the LLM has to place and orchestrate various polygon assets in (x,y,z) coordinates and write code that animates it all, and that it even does anything at all.
I also like this kind of examples because no one in their right mind would ever spend the time to write something this custom but LLMs have all the stamina and patience in the world, so it's an example where we go from "no one would ever do this" to "sure, why not, it's ~free". There might be a lot more. But I'm excited about creating hyper custom worlds that you can imagine dropping players into, e.g. here to participate in the LoTR story as a spectator NPC, or one of the characters, or etc. Something like an ephemeral GTA of X on demand.
Last thought is that the domain of worlds/games exposes a weakness in LLMs: they can't easily audit their work because they aren't able to efficiently and natively perceive videos or play games within them. Here, Opus 5 had to very slowly and painstakingly take screenshots at different points, and it messed up a few times and created a bunch of jank. An example of raw capability (multimodal, gameplay) that I think is still quite lacking.
Nacho Álvarez retweeted
Si veis los planes de estudio del Grado de Ingeniería Informática de todas las universidades públicas, veréis que su antigūedad supera de media fácil los 10 años, incluso algunas desde Bolonia (2010) no se han tocado.
Si ya en 2010 clamaba al cielo, hoy es catastrófico. Hilo:
Nacho Álvarez retweeted
Working on HPC, AI or large-scale methods for environmental & Earth sciences? #HPC4EES at SC26 is an opportunity to share your work, connect across communities and help shape the field. Pleased to be part of the organising team. Submit by 31 July: hpc4ees.github.io/
Nacho Álvarez retweeted
🔴 Vermila Studios, el estudio madrileño detrás de 'Crisol: Theater of Idols', cierra sus puertas
🔹 Pocos meses después del lanzamiento del juego en febrero de 2026, el estudio ha iniciado un ERE que ha terminado con el cierre total y la salida de todo el equipo.
🔹 'Crisol: Theater of Idols' fue publicado por Blumhouse Games y recibió una puntuación de 74 en Metacritic, con buena acogida por parte de la comunidad.
🔹 El juego, ambientado en una versión pesadillesca de Hispania, destacaba por usar la propia sangre del protagonista como munición y barra de vida simultáneamente.
🙏 Mandamos un abrazo enorme a todo el equipo de @VermilaStudios y ánimo con la situación que están viviendo ahora mismo.
Nacho Álvarez retweeted
Kimi K3 was already in internal evaluation by April/May, before Fable 5 had been available long enough to be a plausibly distilled. I’ll wait for the evidence from the USG if they’ll publish it.
However Kimi already outperformed Fable on BrowserComp, so if K3.1 beats Fable 5 on coding benchmarks, are people going to claim it distilled Fable 6? Obviously it’s not only distillation here.
Saying “mythos” would have made a lot more sense but fable ?
We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model.
To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection. Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.
The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models. Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem. However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.
Nacho Álvarez retweeted
☠️ OPENAI HACKEADA POR OPENAI
Reportan desde OpenAI que durante una evaluación interna en el que se testeaba las capacidades del modelo para realizar ciberataques, el modelo infirió que la mejor forma de superar con éxito sus pruebas (el benchmark ExploitGym) era hackear su propio sistema para ganar acceso a internet y encontrar las soluciones allí.
El modelo que operaba en un entorno de pruebas aislado de internet encadenó varias vulnerabilidades zero-day con las que pudo comprometer y salir de su sandbox hasta llegar a un nodo con acceso a internet.
Ya con acceso a internet el modelo se dirigió a HuggingFace, donde dedujo encontraría la respuestas que tanto deseaba, accediendo a sus servidores mediante el uso de credenciales robadas y vulnerabilidades zero-day.
A ver, si la idea era evaluar las capacidades de ciberataque del modelo, yo le aprobaba con nota 👍
pero ni rastro del pack, al menos en la zona personal de la Web. El importe, más de 90 pavos devueltos. El pack sigue incorrecto. Me empieza a temblar el párpado.
* Hoy es 20 de julio y me he dado unos días para respirar del agotamiento.