Pinned Tweet
El 22 de julio, a las 13:02, una chispa de una máquina.
Cuatro días después, el mayor incendio forestal de la historia de España.
Y en El Valle de Iruelas y El Valle del Tiétar, una frase que se repite casa por casa: "nos hemos sentido abandonados".
🧵
#ifburgohondo
Germán Huertas retweeted
Haters were right: it makes zero sense to burn LLM tokens animating humanoids
instead of draining my Opus 5.5 or Astra limits, i switched to Mixamo and got way smoother zombies in 10 minutes
exact pipeline to copy:
1 - make a mesh on Tripo or Meshy (both have free tokens)
2 -don't rig it, just download the raw mesh
3 - send the mesh to Sol / Astra / Argon / Opus / Sonnet / DeepSeek / Grok
4- ask the LLM to set up Mixamo with your model (pick manually or use an existing pack)
in a couple of minutes = ready-to-use rigged and animated 3D model
Germán Huertas retweeted
分享个我最近给幼儿园做的 fde 案例:
基于 deepseek harness 给幼儿园做了个智能体,分析孩子们的日常生活视频,然后每周出个每个孩子的视频集锦,家长很开心。
然后做了蒙氏教育的 skill,不定期分析孩子成长情况,出个报告,老师们也很开心。
Le pedí anoche a Claude Opus 5.5 Ultracode que auditara los controles de mi juego de padel....
Lanzo + de 50 agentes y ha quemado más de 15mill de tokens...
Esta enfermo.
Germán Huertas retweeted
PS5 rodando XBOX GAMEPASS agora virou o MELHOR CONSOLE DO MUNDO.
BY @ElnTskX3
Germán Huertas retweeted
Opus 5.5 found a pretty clever way to turn concept art into a playable 3D world
The workflow starts with a cel-shaded village concept generated with GPT.
Then:
> Opus 5.5 recreates the layout with simple 3D blocks
takes a screenshot from a fixed camera
> GPT paints over those blocks using the concept art as reference
> Opus builds a shader that projects the painting back onto the 3D geometry
The result still looks like the original illustration, but it’s now a real 3D space you can walk around and enter.
The same pipeline was repeated for interiors.
Characters were generated through GPT Images and then animated with Minimax H3.
This feels like a really interesting shortcut for getting a very specific 2D art direction into an actual playable game without manually modeling every detail.
Me he pulido todo el plan Max de Claude, el Plan Pro de Antigravity y uno Lite de GLM 5.3 que tenía
Ahora tengo ansiedad porque no tengo Tokens disponibles hasta el martes a las 18:00
He estado a punto de comprar Grokbot
Tengo mono
Soy un yonki de la IA?
Germán Huertas retweeted
attention google ai pro subscribers: you can now build with gemini 4 argon in ai studio build / playground 🛠️
get started here: ai.studio/build
Germán Huertas retweeted
Skate 3 is now FULLY PLAYABLE in the browser!🚨
I'm averaging 360fps LOCKED with no graphics downscaling at all. This technology is magic.
It was created by an 18 year old Cyber Security Student over the course of a few days. It runs on any device - mobile, tablet, pc, etc.
Also it has multiplayer support🤯
I would have played this in school all day everyday with my friends back in the day...
No paro de ver ads de @tripoai
Podían donarme unos creditos para mi próximo juego y no me importaría contarlo en mis rrss🙏🏼🤐
Ayer publique un pequeño tweet para desahogarme un poco y alguien se ha tomado el tiempo para darme un feedback honesto sobre alguno de mis proyectos.
Realmente esto es lo que amo de X
Replying to @andupoto @Nocodeboy
Ok, so for 2:
1. Change the name to something that makes it obvious what it does.
2. Drop all those animations, they're pointless.
3. The design doesn't make it obvious what it does. I would fix that.
But honestly, before spending more time on it, you need to ask yourself: does this get better as AI models get better, or it gets irrelevant?
Because it's the kind of thing that with AI models improving, people will just prompt it in chatgpt or other ai generation models.
So what's your differentiator there? Do you automate thumbnails? Do you A/B test them? Do you offer localisation? Text editing tools? Likeness trained on the user that does plenty thumbs?
If you just generate it, which they can do in any AI tool, than you've got no real moat.


























