JJBetancourt retweeted
Se los vengo diciendo desde hace rato. Usen @MulticaAI
Agent platform of 2027.
This is the end game. This is the blueprint for the one and only application layer software that will exist. Every other software will effectively become databases for this layer, the everything agentic operating system for the million agent swarm to work on top of.
This is obviously the end game vision for the application layer of ChatGPT and Claude. I have reasons to believe that an independent company will own this product layer. But most ADE, "enterprise agent platform", or vertical saas companies, are focused on the wrong things far too narrowly, not technically competent, and do not have the product / technical vision to see this inevitable future.
The correct agent platform that wins will function like the base of a garden, from which customers will plant their unique "seeds" to grow their own plants and trees, managed and evolved by agents, humans, and signals from the sensors of the business. This presents a clearer picture than "factory of a factory" imo.
JJBetancourt retweeted
From a filmmaker’s perspective, this is world-class propaganda.
The wardrobe, 16mm film, 4:3 framing. Every creative decision is engineered to make Anthropic feel deeply human, authentic, and mission-driven.
Incredibly effective brand building.
JJBetancourt retweeted
Como no tenían moto le hicieron el homenaje con una bordeadora
finalista #CirujaAwards2026 indiscutido
JJBetancourt retweeted
Tesla es una empresa de software más que una de hardware. El auto que hacen está ok, pero el software de conducción autónoma es otro nivel. En teoría esa nueva versión es 25 veces más segura que la anterior. Mapea en tiempo real todo el tráfico.
JJBetancourt retweeted
Messi entro en modo dark y le tiro un pelotazo en la cara a otro jugador
Por algo juega en el inter vice city CF
JJBetancourt retweeted
Tenía razón. Miren lo que esa interfaz. Pensé que era el GTA 6, pero es un render en tiempo real de lo que ven las ocho cámaras.
Hace tiempo la comunidad espera la llegada de FSD V15 porque la se prometía que era la definitiva (la que te permite dormir al volante). Por fin llegó.
This update is HUGE 👀
- Inkling 2 Mini is now supported in Strata 🥳
- 11,306 tokens/s on prefill and 4,087 tokens/s on output with RTX 5090 😨
- Qwen3.8-Flash-Next receives a 40% performance increase across all hardware we support 🥶
See more at: github.com/Niko1221/Strata/r…
JJBetancourt retweeted
Meet Inkling-2-Mini-Uncensored-GGUF: a lightweight multimodal model that handles both images and text. It's uncensored, GGUF-ready, and built for local inference with llama.cpp. Perfect for devs who want freedom and speed.
NEW: Inkling 2 Mini is the most efficient model we've tested:
- #1 on efficiency (cost vs. intelligence)
- #9 on intelligence (78.4) overall
- #1 on intelligence among open-weights models
- Lowest cost per correct answer ($25.80)
- Biggest accuracy gain from adding web search
It uses a looped transformer, which runs its middle layers three times for every token. That lets a 26B-parameter model reason more without generating more tokens, which keeps costs low. A model that size can't store as many facts, so it relies on search to fill the gaps, and that's where it gains the most.
JJBetancourt retweeted
This is a Saudi Prince's Local AI setup
4 x 512GB M3 Ultra Mac Studios clustered with RDMA over Thunderbolt 5 running @exolabs
JJBetancourt retweeted
MINIMAX H3 KEEPS BLURRING THE DETAILS. THIS 74 MB FILE FIXES IT 😳
one tiny anatomy lora for the minimax h3 video model, text-to-video and image-to-video
- file HMPussyV2.safetensors, 74 mb
- 66,917 downloads on civitai, 13,335 on this v2 alone
- trained on explicit close-ups, start at strength 1.0
- single file, works on the pruned h3 checkpoint
not another full nsfw pack, just the part the base model fumbles.
try it 👇
huggingface.co/Hearmeman/min…
JJBetancourt retweeted
Estaban entrevistando a paredes despues de ganar la sudamericana y se desmayo en vivo
JJBetancourt retweeted
La AI llegó exactamente al punto que quería que llegue, poder hacer vibehardware y que salga realmente bien.
Astra 6.2, una impresora 3D y el límite es el cielo.
Les presento a Mimo, un robot de escritorio hecho en un prompt con GPT Astra 6.2 y mandado a imprimir tal cual me dio los STLs.
La pantalla es una TFT de 2.8” con un ESP32 integrado y los brazos 2 servomotores SG90.
Ya podemos crear nuestros propios juguetes, robots, lo que queramos.
JJBetancourt retweeted
We're grateful for our research partner working with us to study looped transformers.
Goodfire explored how layers pass information to future loops, and which interpretability tools allow us to understand the model's inner workings.
Looped models are weird. A layer can write to itself or read from "future" layers.
@thinkymachines gave us early access to Inkling 2 Mini so we could study its inner mechanisms. We uncover how it coordinates information flow across layers, blocks, and loops.
Studying Inkling 2 Mini gave us cause for optimism: as transformers continue to evolve, we believe our ability to understand their internal mechanisms can evolve with them.
Read the full post: goodfire.com/research/interp…
JJBetancourt retweeted
Inkling 2 Mini is a looped transformer that runs on a laptop.
Looping lets a model think more and write less. At this scale, its thought process remains monitorable; this may not stay the case. We’re also sharing a variable-loop research model to enable further study.
Some of you thought that Inkling 2 Small was too big at 594B parameters. We heard you: Inkling 2 Mini is a looped transformer with just 26B parameters, small enough to run on your laptop, powerful enough for agentic work with a score of 20.1 on AA Index.
thinkingmachines.ai/news/ink…




































