El Diego

Perdulario || Quiscalus mexicanus || Todo lo sabemos entre todos.

[El Diego @Kantzakata]

Hice un mapa para el #30DayMapChallenge (que no voy a completar).

Points: localidades con hablantes de alguna lengua indígena en México.

[César Pineda @cesarpinedar]

Defender la “libertad para negar el genocidio” (bajo la máscara de la libertad de expresión) debe ser el pináculo del individualismo atomista que impone el derecho individual a mentir para colaborar directa o indirectamente con un estado genocida.

La defensa del “derecho al negacionismo” se impone a través de la victimización para dejar de hablar de las verdaderas víctimas : las del genocidio en curso. El pueblo de Gaza. Hay indignación por las primeras y no por la segundas.

La contradicción es evidente: quienes defienden la libertad de expresión para el negacionismo tendrían que aceptar la libertad de expresión en protesta contra el negacionismo. Pero solo se defiende la primera y se condena la segunda.

Si el derecho a la mentira genocida se impone por encima del derecho a la existencia de un pueblo, la libertad negativa muestra de forma evidente la tiranía del ego: la libertad propia se desentiende radicalmente del destino y libertad de los demás.

En nombre de la libertad del individuo, se colabora con la opresión y la muerte de todo un pueblo.

[Alex Veremeyenko @alex_verem]

Everyone assumes LLMs are the future of AI.

The permanent foundation. The layer everything else gets built on.

I’m not so sure.

The historical parallel that fits best isn’t the one most people want to hear.

LLMs are Edison’s DC power grid:
→ Genuinely revolutionary
→ Commercially dominant
→ Solving real problems right now
→ But architecturally limited in ways that can’t be patched

Right domain. Wrong architecture. And the evidence is already here.

Hallucination isn’t a bug. It’s the architecture.

Researchers have formally proven that LLMs cannot learn all computable functions and will therefore inevitably hallucinate when used as general problem solvers.

That’s not a training data problem. That’s math.

A separate paper demonstrated that hallucinations stem from the fundamental mathematical and logical structure of LLMs, making it impossible to eliminate them through architectural improvements, dataset enhancements, or fact-checking mechanisms.

And here’s the part that really gets you:
There’s a direct link between hallucination and creativity in LLMs.

It may be impossible to eliminate hallucination without impairing the model’s most crucial capabilities.

→ The thing that makes LLMs creative is the same thing that makes them lie
→ Fix one, you break the other
→ That’s not a tradeoff you engineer away. That’s a design constraint.

DC power had the exact same structural problem. It couldn’t transmit electricity over long distances.

Not because the engineering was bad. Because the physics made it impossible.

You needed AC. A fundamentally different approach.

The “AC power” of AI is already being built. And it has names.

This isn’t theoretical. People are already building the replacement architectures.

Yann LeCun left Meta and raised $1 billion to prove LLMs are a dead end.

AMI Labs raised $1.03 billion in seed funding at a $3.5 billion valuation in March 2026, making it the largest seed round in European history.

His thesis is simple: LLMs predict the next word. That’s not intelligence. That’s autocomplete at scale.

His core technology, JEPA, operates in latent space, learning abstract representations of reality rather than surface patterns.

LeCun used a vivid analogy: using an LLM to understand the real world is like teaching someone to drive by just talking.

A Turing Award winner didn’t just write a paper about it. He quit his job and bet a billion dollars on it.

Mamba is proving transformers aren’t the only game in town.

Mamba achieves 5x higher throughput than Transformers with linear scaling in sequence length.

Thanks to intensive research in 2023-2025, non-transformer architectures have reached parity with Transformers on key language benchmarks, and in some cases surpassed them.

Hybrid architectures are already shipping.

By 2026, models built on hybrid transformer-SSM architectures can ingest hundreds of pages of text at once, far beyond vanilla GPT-3 or GPT-4.

The alternatives aren’t coming. They’re here.

Meanwhile, look at what the industry is building to keep LLMs functional:

→ Agents (because the model can’t verify its own outputs)
→ Tool use (because the model can’t interact with the real world)
→ Reasoning chains (because the model can’t reason natively)
→ RAG (because the model can’t reliably recall facts)

These aren’t features. These are workarounds.

When you need that many patches, you’re running longer DC power lines and wondering why the voltage keeps dropping.

Now the part everyone actually needs: which skills survive the transition?

When DC shifted to AC, some electrical engineers thrived and some went extinct.

The ones who thrived understood circuits, load management, and power distribution at a fundamental level. Those principles worked on any architecture.

The ones who didn’t? They only knew DC-specific wiring.

The same split is coming. And it’s coming faster than people think.

Here are the skills that transfer no matter what replaces transformers:

�� Systems thinking for AI workflows. Breaking complex tasks into steps an AI can execute. This works whether the AI is a transformer, an SSM, JEPA, or something we haven’t built yet. Architectures change. The need for structured task decomposition doesn’t.

→ Evaluation and verification. Knowing if AI output is right. LLMs have a “Self-Correction Blind Spot” where they can recognize errors but lack the reasoning pathways to correct them.  Whatever comes next will still need humans who can evaluate quality. This skill gets MORE valuable, not less.

→ Data literacy. Understanding what data an AI needs, how to structure it, what’s clean vs. noisy. Every AI architecture runs on data. Past, present, future. The people who understand data will always have leverage.

→ AI-augmented workflow design. Not “how to write a good prompt” but “how to redesign a business process so AI handles the right parts and humans handle the right parts.” This is architecture-agnostic. It transfers to anything.

→ Domain expertise + AI fluency. The most powerful combination is stacking AI fluency on top of deep domain expertise.  A lawyer who understands AI beats a prompt engineer who doesn’t understand law. Every time. Regardless of what model they’re using.

→ Clear problem definition. Prompt engineering is just one implementation of a deeper skill: translating human intent into machine-executable instructions. Whether that instruction is a prompt, an API call, a config file, or something that doesn’t exist yet, the ability to define what you want is permanent.

And here’s what DOESN’T transfer:
→ Memorizing specific model behaviors (“Claude does X, GPT does Y”)
→ Platform-specific tricks that only work on one tool
→ Building your identity around a single product name
→ “Prompt engineer” as a job title instead of a thinking skill
The difference is simple:
→ Transferable skills = understanding WHY something works
→ Non-transferable skills = memorizing HOW a specific tool works

WHY survives paradigm shifts. HOW doesn’t.

The bottom line
The principle behind LLMs is permanent. The architecture probably isn’t.

That’s not bearish on AI. That’s the most bullish take possible. It means the best is still ahead of us.

Use LLMs hard right now. Build with them. Ship on them.

But build your skills around the PRINCIPLES, not the PRODUCTS:
→ Learn systems thinking, not just prompting
→ Learn evaluation, not just generation
→ Learn data literacy, not just tool literacy
→ Learn workflow design, not just model tricks
→ Stack domain expertise on top of AI fluency

The people who do this will thrive in the transformer era AND whatever comes after it.

Edison built a working power grid that lit up Manhattan. It was real, valuable, and changed the world.

AC still replaced it.

[Krasnaiaesgorria @Krasnaiaesrojo]

Un programador egipcio ha creado una página dedicada a cada víctima del genocidio palestino. Hasta ahora ha registrado 72 000 nombres. Cada punto de luz representa una víctima y, al pulsarlo, verás su nombre y fecha de nacimiento. Puedes filtrar por edad.
https://t.co/1dCihg3ixC

[Instituto Mora @institutomora]

I Taller Introducción a las herramientas cartográficas para la investigación histórica.
🔹Curso presencial.
🗓️Del 21 de abril al 12 de mayo de 2026.

Inicio de cursos: 21 de abril.
Inversión: $900.00 MXN.
Correo: [email protected]

[Python Developer @PythonDvz]

How to create interactive maps in Python with folium

[Cristina Rivera Garza @criveragarza]

Con base en papeles que todos pueden consultar en el Archivo Histórico de la Secretaría de Recursos Hidráulicos, este libro es mi Rulfo mío de mí. Hay viaje, documentos, ficción, caminatas—y tiene un último capítulo en Mixe. No cuenta con la aprobación de la Fundación Rulfo :)

[Yuri Knorozov @YuriKnorozov]

Knorozov on a 2022 stamp of Russia #LinguistsOnStamps

[freeCodeCamp.org @freeCodeCamp]

AI tools can help you edit images - from colorizing black and white photos to upscaling and denoising.

In this guide, Manish teaches you some core techniques like GAN enhancement, super-resolution, and artifact removal.

You'll learn when to use each, and how you can plug them into your apps using APIs.

https://t.co/L969ifp52A

[UNAM @UNAM_MX]

#BoletínUNAM Gastronomía mexicana, deliciosa riqueza > https://t.co/GTSK3N92mf

[MappingGIS 🚀🌍 @MappingGIS]

Overture maps: la mejor cartografía vectorial global https://t.co/sM5ArqbT4Q #datos #googlemaps #OpenStreetMap

[El Diego @Kantzakata]

🤖 Taller fundamentos para la transcripción automatizada de audios.

📅⌨️ Fechas: del 26 al 30 de enero.
⏰ Horario: 15:00-18:00 (GTM-6)
🐍 Se emplearán conocimientos básicos de Python
⌨️ Liga al formulario de registro:
https://t.co/9NHGRHuH8K

[PUIC UNAM @puic_unam]

📚✨ La voz de #RosarioCastellanos multiplicada en 18 lenguas encontró su eco en la #FILGuadalajara

Gracias por ser parte de este acto de justicia poética y lingüística. ¡Seguimos trabajando por la #DiversidadCultural!

@FILGuadalajara @Unicach_Press @redaltexto @librosunam

[Literatura y Fomento a la Lectura UNAM @LiteraturaUNAM]

Lamentamos el fallecimiento de la excepcional Dra. Margit Frenk (1925-2025).

Profesora emérita, investigadora incansable y maestra de generaciones, dedicó su vida a rescatar y comprender la riqueza de la lengua española, desde la lírica popular hasta el Siglo de Oro. Su trayectoria en la UNAM, El Colegio de México y la Academia Mexicana de la Lengua deja un legado académico fundamental.

Descanse en paz. 🖤

[Foto: Revista Común]

[Filológicas @IIFL_UNAM]

📚✨ La palabra no solo comunica: también construye vínculos y responsabilidades.
Acompáñanos mañana en la conferencia “Los cuidados en la comunicación y la responsabilidad al hablar”, a cargo de Jeanett Reynoso Noverón.

[César Pineda @cesarpinedar]

Los países que reconocen a Palestina son mayoría en el mundo.
Confundir el Estado de Palestina (que ha reconocido apenas hace unos días Canadá) con la Palestina histórica (se habla de Palestina desde la antigüedad), es sólo una narrativa engañabobos, para confundir el debate internacional: Israel ocupa ilegalmente tierras palestinas, (resolución 452 de 1979; resolución 478 de 1980; 58/292 de 2004 de la ONU, entre varias más) y hoy comete genocidio (ver link abajo)
No hay confusión aunque las derechas quieran negarlo: Israel ocupa militar e ilegalmente tierras palestinas (que es el origen de la cuestión Palestina) y hoy comete genocidio como lo afirma la ONU, Amnistía Internacional y la relatora Especial para los Territorios palestinos, Francesca Albanese.
Israel es responsable de ocupación, apartheid y genocidio. 🧵de pruebas

[INAH @INAHmx]

¡Asiste al Coloquio "Los Colegios de Propaganda Fide en América, Filipinas y España (Siglos XVII al XX)"!

🗓️ Del 23 al 25 de julio, 10 a 14 h.
In memoriam fray Francisco Morales | OFM

Transmisión por el canal de la @cnan_inah: https://t.co/bMyb8JhMHu

[Biblioteca de Investigación Juan de Córdova-FAHHO @BibliotecaJdeC]

Cierre de la décima edición del #COLOV. Agradecemos a los participantes,instituciones, colaboradores, voluntarios y personal que hizo posible este evento. Hasta el próximo encuentro de investigadores, promotores y hablantes de las lenguas otomangues y vecinas! #BIJC #FAHHO #UNAM

[Biblioteca de Investigación Juan de Córdova-FAHHO @BibliotecaJdeC]

Inauguración de la décima edición del Congreso de lenguas otomangues y vecinas #COLOV, dedicado a la memoria de la lingüista Kathryn Josserand y su trabajo con la dialectología del mixteco. Agradecemos al personal, voluntarios y servicio social que hacen posible este evento.