Rubén

A little update on how Audacity 4 is progressing. We've just about completed all the core editing work now and I wanted to demonstrate some of the workflow improvements we've included. Let me know if you'd like more little updates like this.

@www.error500.net qué decías que los que montan IDE con IA iban a sacar sus modelos? XD x.com/windsurf_ai/...

Much ado about the role of renewable energy (RE) in the Spanish blackout. The investigation into the effective cause is still ongoing, but many pointed to the fact that RE has no ‘inertia’. Is that true? The answer is: yes, but ... 🧵

Países con más porcentaje de renovables que España en la Red eléctrica: España: 56% Alemania: 60% Dinamarca: 81% Austria: 87% Suecia: 87% Noruega: 96% Uruguay: 98% Apagones generales: CERO A los oportunistas anti-renovables que aprovechan el río revuelto para difundir bulos se les frena con DATOS.

Qué random que puede ser internet <3 no necesitamos algoritmos conductistas para disfrutarla

Jiji, jaja, pero con Nintendo no hay huevos eh Sam? Captura al vuelo pq OpenAi sí bloquea a Pokémon, no sea que…

Can a Large Language Model (LLM) with zero Pokémon-specific training achieve expert-level performance in competitive Pokémon battles? Introducing PokéChamp, our minimax LLM agent that reaches top 30%-10% human-level Elo on Pokémon Showdown! New paper on arXiv and code on github!

Such a big fan of Dan Hollick's work (he doesn't appear to be here? Please let me know if he is and I searched wrong.) He has such an instantly recognisable aesthetic with his work. I aspire to this level of brand recognition. www.makingsoftware.co

Doom now runs in @typescriptlang.org types. What a journey this one's been.

Today, we’re launching new algorithms that enable 10x longer context lengths & 90% less VRAM for training Reasoning Models (GRPO). Using Unsloth, you can now train your own reasoning model with just 5GB VRAM for Qwen2.5-1.5B with no accuracy loss. Blog: unsloth.ai/blog/grpo

If you’ve not heard of these, they’re extremely cool tech! evanw.github.io/thumbhash/ blurha.sh

@www.error500.net te animas? huggingface.co/learn/agents...

HTML, CSS, JS & A11y course stack: - HTML For People by @blakewatson.com: htmlforpeople.com - Complete CSS by @piccalil.li: complete-css.com - Joy For JavaScript Developers by @joshwcomeau.com: www.joyforjs.com - Practical Accessibility by @sarasoueidan.com: practical-accessibility.today

Suspension Bridge xkcd.com/3048

Reading timdellinger.substack.com/p/hey-wait-i... has me thinking about “performance” in employees more generally. A mistake that irks me (which this article does not make) is when people speak about “high performers” and “low performers” as some mostly-static state of being.

🎉🎉 New release of my #dashboard of #opensource #lowcode tools modeling-languages.com/dashboard-of... With new insights regarding the % of low-code tools that also use #AI, sell themselves as #low-code or use #models

@www.error500.net y si lanzas 3, 5, 10 Deep Researchs y luego lo pasas por un LLM evaluador? No subirias la fiabilidad? Aún no estamos ahí computacionalmente o en precio, pero solo es un horizonte muy cercano y seguro que hay clientes VIP que tienen eso

New year, new blog post: I had a random question, what happens when LLMs are prompted to write better code, again and again? Do they actually write better code? The answer is yes*! minimaxir.com/2025/01/writ...

La silueta de la cúpula destruida por el fuego es un icono de la divulgación científica en nuestro país. El Planetario de Pamplona debe renacer de sus cenizas.

I couldn't find any good games for my password manager. So I made one. 1Crossword connects to your 1Password vault and generates a crossword entirely out of your passwords! The crosswords are fun, simple, and great for sharing on social media when you finish. Enjoy!

Radon xkcd.com/3037

It's that time of year again to brush up on your regular expressions. Do you dare to discover the horrors that await in... The Typing of the RegEX thetypingoftheregex.com

It's alive! Input layer at the top, output layer at the bottom. Connection colour represents positive (green) or negative (orange) weight, thickness and opacity represents how large that weight is. Really neat to see the training process strengthening specific connections in the network.

General consensus in the replies and quotes of this seems to be that the entire concept of "AI skills" is a joke - how hard is typing text into a chatbot, really? I will continue to argue that it's genuinely difficult, and that the challenge in using these tools is widely underestimated

La vida es lo que pasa entre barra de progreso y barra de progreso del AI Agent que programa por ti xD

Em van convidar a parlar d'educació financera, aquest n'és el resultat youtu.be/P4Z2MugqdqI?...

#HolaCiencia! Un nuevo riesgo asociado a las #megaconstelaciones de #satélites en órbita baja como #Starlink: la contaminación directa de la estratosfera con litio, aluminio, cobre y otros metales debido a su vaporización durante la reentrada • #FotoCiencia Paul Fleet 🛰🛰🛰🛰

¡#HolaCiencia! El área de un #triángulo es la mitad del producto de su base por su altura • #VideoCiencia por Magic PI #PruebaVisual

Átomo de deuterio con loros • #HumorCiencia vía j4cks0nB en Reddit

By far the best coverage of o3 is this essay by François Chollet, it's crammed with interesting insights beyond just reporting on the benchmark score: arcprize.org/blog/oai-o3-... Published my own notes on that here: simonwillison.net/2024/Dec/20/...

@www.error500.net Esta charla te va a encantar, no te asustes porque presenten al ponente en euskera, la charla es en español www.youtube.com/watch?v=dzeR...

i'm building an experimental tool for exploring 25 years of my old sketchbooks, with image and text recognition powered by gemini

Zetatesters is BACK! zetatesters.com/2024/12/zt-1...

I can now run a GPT-4 class model on my laptop (The exact same laptop that could just about run a GPT-3 class model 20 months ago) The new Llama 3.3 70B is a striking example of the huge efficiency gains we've seen in the last two years simonwillison.net/2024/Dec/9/l...

This demo of structured data extraction running on an LLM that executes entirely in the browser (Chrome only for the moment since it uses WebGPU) is amazing My notes here: simonwillison.net/2024/Nov/29/...

As a father I want the same xD

Neutrino Modem xkcd.com/3017

www.elotrolado.net/noticias/tec... @scanbcn.bsky.social

If you want to apply Gen ai in education, Check the Safe AI Manifesto manifesto.safeaieducation.org

Recursion (Source.)

dev.to/hatem_zidi/i...

La puta experiencia open.substack.com/pub/divagati... #gastronomia #experienciagastro #restaurantes