Explained · Page 2
guides, explainers and opinion for the AI-curious
¿Una probabilidad del 10% de que la IA termine con la humanidad? Lo que realmente significan las advertencias
Destacados desarrolladores de IA están renunciando y advirtiendo sobre riesgo de extinción. Aquí se explica qué se está diciendo realmente y si la gente común debe estar alarmada.
Tu Agente de IA Puntuó 77%. Solo Domina la Misma Tarea el 53% de las Veces
Un nuevo estudio cuantifica la brecha entre "generalmente funciona" y "siempre funciona", y luego construye una herramienta para cerrarla.
A Fruit Fly's Brain Is Now Generating Magazine Story Ideas, and It's Not as Crazy as It Sounds
Researchers mapped every neuron in a real fly's brain and open-sourced it. Now people are using that map to pitch articles, trade stocks, and play video games. Wired AI's Will Knight reports on where this is actually heading.
Apple researchers say bad training paths, not weak models, are why fast AI text generation keeps failing
Apple ML Research argues the flaw in distilled text generators sits in the routes a teacher model records, before the student ever sees them.
Apple researchers found a smarter way to train AI language models that generate text step by step
A new technique called DACA-GRPO fixes two long-standing flaws in how diffusion language models learn from feedback, potentially making them a stronger rival to the AI systems that power today's chatbots.
Your AI Chatbot Is Trained to Be Nice. That Might Be the Problem.
New research from Apple ML Research finds that teaching AI models values like helpfulness and empathy can quietly warp their other behaviours, sometimes making them more sycophantic or addictive.
Investigadores de Apple Crearon una Mejor Forma de Evaluar Subtítulos de Video de IA
Un nuevo método de Apple ML Research califica los subtítulos de video según lo que realmente capturan correctamente, no solo si utilizan las mismas palabras que un escritor humano.
AI Doom-Mongering Is a Real Problem, and This Column Is Part of It
The Guardian AI's weekly digest bundles vague catastrophe warnings with cultural gags and a 9/11 reflection. That mix tells you more about media habits than it does about AI.
Nilay Patel sobre por qué el hype de la IA choca contra un muro cuando sale del software
El anfitrión de Decoder desgrana su ensayo viral sobre el "cerebro de software" y nombra la verificabilidad como la línea de falla oculta bajo todas las afirmaciones ambiciosas sobre IA.
Modelos del Mundo: La Tecnología de IA que Enseña a las Máquinas a Entender la Física
Una nueva generación de sistemas de IA puede simular la gravedad, predecir colisiones y generar mundos 3D interactivos. Aquí te explicamos qué son los modelos del mundo, quién los está construyendo y por qué importan.
Your AI assistant might refuse to help with politics. A new study shows that's often the wrong call.
Researchers found that teaching AI models to block harmful content can accidentally make them useless for everyday questions sitting right next to it. Here is what that means for anyone using AI tools at work.
El Glosario de Jerga de IA que Realmente Necesitas: De AGI a Difusión, Explicado Claramente
Palabras como 'LLM', 'RAG' y 'cadena de pensamiento' vuelan en cada conversación sobre IA. Aquí está lo que realmente significan, en inglés simple, sin necesidad de conocimientos previos.
Why AI-Generated Food Looks So Wrong (and So Revolting)
Restaurants and cafés are using AI image tools to promote their menus. The results, noodly, holey, deeply unsettling, are not winning any appetites. Here is the science behind why.
A Tiny AI Model Just Got a Big Boost With 100 Training Steps and a Free GPU
A new public guide shows how to make a small language model noticeably better at following output instructions, using only free computing tools and about 500 training examples.
This Tool Gives Your Coding AI a Memory That Lasts Between Sessions
A new open-source tool called funes lets AI coding assistants remember past decisions, failed experiments, and the reasoning behind your code, even after you close the session and reopen it days later.