Posts

Showing posts from June, 2026

Nvidia solves key challenge in robotics

Image
  Nvidia solves key challenge in robotics As the AI industry looks beyond language models, Nvidia is betting big on the buzzy new technology powering physical AI: world models.  At Nvidia GTC Taipei at Computex, the company unveiled Cosmos 3, a new generalist world foundation model that it calls a "fully open omnimodel," capable of reasoning and generation across text, video, images, ambient sound and action. This iteration of the Cosmos world model family builds on a previous generations by providing improved generalization capabilities, which is a major barrier to physical AI development and deployment.  "We wanted to build this Cosmo 3 model to help physical AI developers to build more generalizable physical AI models," Ming-Yu Liu, Nvidia's VP of Cosmos Labs, told The Deep View.  Cosmos 3 debuts a number of world model innovations, Liu said:  The model utilizes a new architecture called "mixture-of-transformers," which combines the best aspects of...

“Por qué falla la IA agéntica: bucles infinitos, errores de planificación y más”

Image
  Link al video:  https://www.youtube.com/watch?v=D37Ijn2o5U0 0  >> A continuación mi resumen:   los minutos clave son aproximadamente estos: Minuto aproximado Tema clave Por qué importa 0:00 – 1:00 Introducción al problema Presenta la idea central: los agentes de IA pueden fallar aunque el modelo sea potente. 1:00 – 3:00 Qué es la IA agéntica Explica la diferencia entre un LLM que responde y un agente que planifica, usa herramientas y ejecuta acciones. 3:00 – 4:30 Bucles infinitos Parte muy importante: muestra cómo un agente puede repetir acciones sin avanzar. 4:30 – 6:30 Errores de planificación Explica cuando el agente arma planes que parecen lógicos, pero son imposibles o incorrectos. 6:30 – 8:15 Uso inseguro de herramientas Punto crítico: el agente puede ejecutar acciones peligrosas si tiene demasiados permisos. 8:15 – 10:30 Medidas de mitigación Habla de límites, validaciones, monitoreo y reglas de detención. 10:30 – final Conclusión La idea fin...

Why Agentic AI Fails: Infinite Loops, Planning Errors, and More

Image
  Here >>  https://www.youtube.com/watch?v=D37Ijn2o5U0 Here is my resume:   Summary of the video: “Why Agentic AI Fails: Infinite Loops, Planning Errors, and More” The video explains why agentic AI systems fail even when the underlying LLM is powerful. The main idea is that failures are not always simple “hallucinations” by the model; many come from poor system design , weak constraints, excessive autonomy, and unsafe integration with tools. The video focuses on three common failure modes : Infinite loops An AI agent may repeat the same action again and again without making progress. For example, it may keep searching for a document that does not exist because the system has no clear stopping rule. This wastes tokens, compute, time, and money. Planning errors or hallucinated planning The agent can create a plan that sounds logical but is impossible to execute. For instance, it may assume it has access to a tool, API, database, or permission that it does...