Horacio Pérez Sánchez
horaciops.bsky.social
Horacio Pérez Sánchez
@horaciops.bsky.social
🧬 Researcher | 🔍 Molecular Discovery & AI |🎙️ Host of "Investig la I." (350+ ep)| ❌ Don’t follow the algorithm |🧪 Model the system |❓ Ask better questions | https://horacio-ps.com
Reposted by Horacio Pérez Sánchez
¿Seguirás escribiendo código línea a línea en 5 años? La programación con IA ya no es solo ChatGPT y copiar/pegar. Es orquestar agentes. Así está cambiando todo. 🧵 1/16
February 1, 2026 at 2:49 PM
¿Seguirás escribiendo código línea a línea en 5 años? La programación con IA ya no es solo ChatGPT y copiar/pegar. Es orquestar agentes. Así está cambiando todo. 🧵 1/16
February 1, 2026 at 2:49 PM
En este episodio parto de algo que, a primera vista, parece no tener nada que ver con la investigación que hago en mi grupo de investigación ni con el día a día de cualquiera de nosotros: los agujeros negros. 🧵 1/10
January 28, 2026 at 8:37 PM
Reposted by Horacio Pérez Sánchez
En la investigación no todo es generar ideas o redactar manuscritos. También hay un momento inevitable en el que otras personas evalúan aquello en lo que hemos invertido mucho tiempo y esfuerzo. 🧵 1/13
January 23, 2026 at 6:02 PM
En la investigación no todo es generar ideas o redactar manuscritos. También hay un momento inevitable en el que otras personas evalúan aquello en lo que hemos invertido mucho tiempo y esfuerzo. 🧵 1/13
January 23, 2026 at 6:02 PM
Will we need new programming languages so that AI can understand us better — or to make "vibe coding" really work? 🧵 1/14
January 20, 2026 at 11:11 PM
Have you ever wondered if the "Ralph Wiggum" technique in vibe coding could actually be improved in ways that nobody has really tested yet? 🧵 1/12
January 12, 2026 at 7:35 PM
¿Y si el mayor cambio en programación no tuviera que ver con aprender un nuevo lenguaje, sino con dejar de programar como lo hemos hecho siempre? Durante décadas, el progreso fue claro y más o menos lineal: mejores máquinas, mejores lenguajes, mejores herramientas. 🧵 1/14
January 10, 2026 at 4:37 PM
We spend hours optimizing tools, apps, workflows. But almost no one talks about something more basic: where you are physically when you think changes what you can think. Not how well you think. What kind of thinking is possible. 🧵 1/9
January 7, 2026 at 10:36 PM
Durante décadas hemos señalado al colesterol como el gran culpable de la aterosclerosis.
Pero ¿y si una parte clave del problema no viniera de la dieta… sino de la microbiota intestinal? 🧵 1/7
January 6, 2026 at 4:32 PM
Reposted by Horacio Pérez Sánchez
Participa en nuestra sección "Artículo del mes" y opta al Premio ‘María Teresa Miras’ al mejor artículo de jóvenes de la SEBBM que entregamos de manera anual durante nuestro congreso SEBBM.

Anímate y participa 😜
https://f.mtr.cool/oneifmxdbf
January 5, 2026 at 10:02 AM
Accelerating enzymatic mechanism studies: from QM/MM to ML/MM The problem
QM/MM methods are the gold standard for in silico studies of enzymatic reaction mechanisms. 🧵 1/12
January 5, 2026 at 8:05 AM
Can we extract knowledge from Large Language Models — or even use them to create scientific data? Large Language Models are often described as having "compressed" enormous amounts of information from the scientific literature. But this raises a fundamental question: 🧵 1/10
January 4, 2026 at 11:25 AM
Organizing research knowledge in Obsidian is harder than it looks. Folders give order, links give meaning. Options I see: Johnny-Decimal folders, flat notes + human MOCs, AI-generated MOCs, or a few human-written core notes with derived views. What do you use to actually think?
January 4, 2026 at 1:26 AM
Reposted by Horacio Pérez Sánchez
@horaciops.bsky.social Mariano González Campo, filósofo, filólogo, antropólogo social y traductor de literatura nórdica. Entrevistado por Horacio Pérez Sánchez, conductor del grupo Investigando la investigación
youtu.be/TmgxFFx5tyo?... #filosofia #filosofiaenabierto #literatura #literaturanordica
371. Pasión por el Norte: Mariano González Campo, traductor de lenguas nórdicas
YouTube video by Investigando la investigación
youtu.be
November 15, 2025 at 5:55 AM
What if our limits are not human, but only the computing limits of the simulation?
When intuition fails, it is not lack of wisdom, but the processor slowing.
Poetry resists collapse: many meanings alive.
Hamster wheels: only loops inside the code.

#PhilosophyFiction #ExistenceLoop #PWL
Bostrom says: maybe we live in a simulation by higher beings.
But what if not?

What if we build it ourselves, together, without knowing?

A closed loop of symbols, language, culture. No outside. Only us, repeating the shared construction we call “world”.

#PWL
September 6, 2025 at 4:22 PM
September 6, 2025 at 3:30 PM
Bostrom says: maybe we live in a simulation by higher beings.
But what if not?

What if we build it ourselves, together, without knowing?

A closed loop of symbols, language, culture. No outside. Only us, repeating the shared construction we call “world”.

#PWL
September 6, 2025 at 3:29 PM
Notes on complex topics often scatter into fragments. I’m testing a “multi-level zoom” system: numbered depths + parallel perspectives. AI then helps compile, linking pieces with color markers. Multiple views before the final shot. Anyone tried similar? #Obsidian #notetaking #PWL
August 27, 2025 at 3:41 PM
Leveraging LLMs for A/B testing means more than content tweaks—it’s a new research paradigm. LLMs could propose hypotheses, test contexts, spot counterexamples, even optimize content in real time. Yet, reliability, interpretability, and rigor remain concerns. Tool or trap? #LLMResearch #PWL
August 27, 2025 at 3:41 PM
Some philosophers think like mathematicians: step by step, patient, precise. Spinoza starts with definitions and ends with God = Nature. Hegel shows conflict drives growth. Parfit finds identity less key than continuity. Sometimes only a long path reveals truth. #Spinoza #Hegel #Parfit #PWL
August 27, 2025 at 3:39 PM
Reposted by Horacio Pérez Sánchez
Charlando en "Investigando la investigación" con @horaciops.bsky.social sobre mi experiencia como divulgador y explorador científico, sobre misiones científicas, globos estratosféricos, documentales como "Into the Aurora" y mucho más.

#Ciencia #Divulgación #tecnologia

youtu.be/ud8jnQs0lHU
343. Fernando Ortuño y las auroras boreales: ciencia y divulgación desde la estratosfera
YouTube video by Investigando la investigación
youtu.be
April 5, 2025 at 11:28 AM
Reposted by Horacio Pérez Sánchez
STELLAR Enables CPU-Based Fragment Docking and Reconstruction of Large Peptides Without AI Assistance chemrxiv.org/engage/...

---
#proteomics #prot-preprint
June 3, 2025 at 11:20 AM
Reposted by Horacio Pérez Sánchez
The further a researcher or inventor moves from their previous work, the less cited their latest work will be

https://go.nature.com/44VjEW9
Shifting research focus comes with the risk of reduced impact
An analysis of millions of scientific papers and patents shows that the further a researcher or inventor moves from their previous work, the less cited their latest work will be.
go.nature.com
May 30, 2025 at 1:14 PM