🚨 ¡Adiós a vLLM y a las GPUs que cuestan un riñón!
El nuevo monstruo open source ya está aquí y funciona en tu portátil con solo 25 GB de RAM… ¡sin GPU! 🤯💻
Se llama Colibrí. Un solo desarrollador italiano lo creó desde cero en C puro, sin dependencias pesadas ni frameworks raros.
El truco: en vez de cargar los 744 mil millones de parámetros del GLM-5.2 (MoE) enteros en memoria, solo carga al instante los experts que necesita desde el SSD. El resto se queda en disco.
Resultado:
✅ ~10 GB de RAM ocupados de verdad
✅ Corre en hardware normal
✅ Código limpio, rápido y minimalista
Esto no es otra herramienta más. Es la prueba de que los modelos frontier ya no tienen que vivir solo en centros de datos carísimos.
El proyecto está empezando y la comunidad ya lo prueba en setups más potentes (mejores SSD = más velocidad).
📌 Guarda este post.
Repo en los comentarios.
¿Lo vas a probar? ¿En qué máquina lo correrás? Comenta abajo 👇
Elon Musk just proved every sighted person on Earth is blind.
Your eye captures 0.0035% of the electromagnetic spectrum.
You are not seeing the universe. You are seeing the sliver your biology decided was enough to keep you alive.
That was never vision. That was a survival filter bolted onto your perception four hundred million years ago.
No one has ever removed it.
Musk: “Blindsight will enable those who have total loss of vision to be able to see again. Including if they have lost their eyes, or the optic nerve.”
No eyes. No nerve. The entire optical system physically absent from the skull.
Neuralink does not rebuild what broke.
It routes around biology entirely. It streams synthetic signal straight into the visual cortex.
Your eye never saw anything.
Your brain did.
Your brain sits in total darkness inside a vault of bone.
It has never seen the sun. It has never seen anything.
It builds reality out of whatever electrical signal the eye allows through.
The eye is the bottleneck between your mind and the universe.
Neuralink removes the bottleneck.
Musk: “Maybe have never seen, were even blind from birth.”
A human who has never perceived a single photon of light.
Given sight for the first time.
Not through medicine. Through engineering.
Musk: “You can see in radar, you can see in infrared, ultraviolet.”
Infrared is pouring off every surface in the room around you.
Radio waves are passing through your body right now.
Ultraviolet is painting patterns across everything you have ever looked at.
You cannot see any of it.
Your biology locked you out before you were born.
Musk: “Superhuman capabilities.”
The person born blind would not be restored to human sight.
They would see more of the universe than any sighted person who has ever lived.
The blind would out-see the sighted.
You will not pity them.
You will envy them.
Musk: “Cybernetic enhancement.”
Every human sense is a sensor converting the world into electrical signal.
Replace the sensor and the brain does not care where the signal came from.
It just processes.
Your senses were never built to show you reality. They were built to show you just enough of it to survive.
Evolution built a keyhole. We called it sight.
Elon Musk is not fixing blindness.
He is fixing sight.
Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.
https://t.co/i8HpU7w64k
Introducing GPT-Live, a new generation of voice models for natural human-AI interaction.
Rolling out in ChatGPT starting today.
You’ll want to turn the sound on for this one.
Meet Dory: a free, open-source Docker Desktop and OrbStack alternative for Mac.
One shared VM for all your containers (~4.7x less idle RAM), a real docker socket, one-click Kubernetes, full Linux machines for all dev work. ~6 MB, no Electron, no accounts, GPL-3.0.
https://t.co/rlMlT1cNau
I gave Fable 5 one job: write custom WebGPU kernels for Gemma 4 inference.
It climbed to 84 tok/s, then hit a wall, insisting further optimization was impossible.
Hours later, Anthropic rolled back invisible LLM development safeguards, and it hit 255 tok/s.
The next day, access to Fable 5 was suspended globally.
Show Codex a workflow once. Reuse it as a skill.
Record & Replay lets you show Codex a recurring task, like filing an expense report or submitting a time-off request.
Codex turns that demo into an inspectable, editable skill.
You control when recording starts and stops.
I love how open the Codex app is, you can run any model you want, even local.
These are the only configs you need to use the app with Gemma 4 via Ollama.