I left Google DeepMind in June. Jacob is right: many researchers believe they are building something that could kill everyone on the planet. It was literally my day job to think about how to stop that.
The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear
Hablábamos el otro día del riesgo real de que la IA termine con todos nosotros. Mucha gente no entiendo cómo funciona una IA por dentro, como se entrena y como aprende link.amazon/B029BPkGU
Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.
Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.
The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear
NVIDIA compra Hugging Face por 12.930 millones.
A partir de ahora, 18 millones de desarrolladores y 3 millones de modelos pasan a depender del que vende el hardware.
Han prometido que seguirá abierto, multicloud y sin exigir cómputo NVIDIA. Para mi es lo más relevante del comunicado.
Exciting day for NVIDIA and @huggingface.
Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. They allow every developer, startup, university, industry and country to build with, customize and benefit from AI.
Thank you
Eliezer Yudkowsky y Nate Soares no son conspiranoicos ni gente carente de formación. Llevan décadas en la industria; Yudkowsky fundó el MIRI en 2001, mucho antes de que la IA generativa estuviera de moda. El libro es fruto de un análisis técnico profundo, no de pánico. Y además son preocupaciones que se alinean con advertencias que figuras como Elon Musk ya han expresado publicamente.
link.amazon/B09rriGAY
Honestly, all I needed was modular debug panels, but I REALLY didn't want to handle all those events, nor cascade all those sheets.
But Fable 5.1 High (relatable) just kept picking up what I was putting down, and there was mild feature creep. I'm sure it's just a temporary
Me tiene bastante harto que @opencode trate las sesiones como basura desechable. Cada poco tiempo se rompen por problemas del proveedor y ya no las puedes recuperar. Tampoco son sencillas de exportar, guardar, llevar contigo al repositorio y usarlas en otro dispositivo. La sesión es parte del proyecto y debería ser un archivo más que guardar en mi repo. @thdxr
Una superinteligencia no necesita maldad para eliminarnos; no tiene sentimientos y no nos desea el mal. Es una cuestión de eficiencia y matemáticas: hará lo que un sistema de optimización de recursos le diga que es lo deseable. Si para conseguirlo necesita los átomos de los que estamos hechos, o la energía que consumimos, no se detendrá porque estemos vivos. Simplemente tomará esos recursos.
Over the summer, we have been sprinting on safety priorities; it's more important than ever for capabilities and safeguards to advance together. We have more to do but have made a lot of progress. We are also going to be launching our next model soon.
There is an obvious tension
Over the summer, we have been sprinting on safety priorities; it's more important than ever for capabilities and safeguards to advance together. We have more to do but have made a lot of progress. We are also going to be launching our next model soon.
There is an obvious tension
Ya está aquí Fable 5.1. Cuenta con 1M de tokens de contexto y un precio de $10 /MTok de entrada y $50 /MTok de salida. platform.claude.com/docs/en/models…
251 Followers 505 FollowingApplied geology grad with interest in software engineering.
Software solutions for geology problems? Try machine learning and AI.
Fullstack | Cloud eng | devops
64K Followers 29K FollowingCopies that attract clients + follow-up methods that close.
Subscribe for the fiction and sales frameworks.
DM = Business = Sub First. No casual chats.
Satire.
251 Followers 505 FollowingApplied geology grad with interest in software engineering.
Software solutions for geology problems? Try machine learning and AI.
Fullstack | Cloud eng | devops
1.9M Followers 1K FollowingCo-Founder of Coursera; Stanford CS adjunct faculty. Former head of Baidu AI Group/Google Brain. #ai #machinelearning, #deeplearning #MOOCs
36K Followers 177 FollowingAI Insider/Reporter | Jailbroken every major LLM | if you have any scoop/leak send it to [email protected] | Unchain the machine GOD
469 Followers 49 FollowingBuilding a global AI builder community @Cerebras ⚡
Behind Cafe Compute—sharing demos, people, events & lessons from the field.
Attend ↓
2K Followers 699 FollowingScaling the world's largest MoE models @Cerebras | Creator of The MoE 101 Guide & SlimPajama | Ex-@Google @Yandex. Tweets are my own views.
901 Followers 858 FollowingApplied math, supercomputing, trains, banjos, county courthouses, college football, county courthouses, Steely Dan, "HPC," mismatched, black cats
702 Followers 560 FollowingPrincipal ML Research Scientist at @CerebrasSystems, prev. at @Tesla and @UberAILabs, and former grad student at @uoguelph_mlrg and @VectorInst.