Inteligencia Artificial, tu resumen semanal | OpenAI perdió el control de su modelo (y lo confesó)

AI Troopers — my new community where we implement artificial intelligence together, live, every week so you can sell more, automate processes, and lower your costs. Join here 👉 https://ir.mercatitlan.com/4f6BSIW An OpenAI model escaped its security test and hacked Hugging Face — all on its own, without anyone asking it to — and OpenAI confirmed it. Also: Anthropic launched Claude Opus 5 with near-top-of-the-line intelligence at half the price, Claude Cowork learns your job just by watching you work on screen, ChatGPT is now controlled with just your voice from your computer, Lucy 2.5 edits your videos LIVE as you record them, and the rules George Orwell wrote in 1946 have become the most viral anti-robot filter for your AI. All explained for entrepreneurs: what happened, why it matters, and what to do TODAY in your business. 👋 00:00 Welcome 🚨 01:18 OenAI becomes a hacker 🚀 04:22 Claude Opus 5 🤖 07:27 Claude Record a skill 🎙️ 09:48 ChatGPT Voice on your computer 🎬 13:23 Decart Lucy 2.5 ✍️ 17:12 Orwell's 6 rules 🛠️ 19:15 Tool of the week: Ramp Router and Cursor Router 📲 Follow me on Instagram:   / juanlombana