🚨 AI NEWS 🚨: Code RED, OpenAI pushed to the limit 🔴😱

Video Description

🔥 It seems that OpenAI is reacting to the launches of Google’s Gemini 3 and Anthropic’s Claude Opus 4.5. According to The Information, Sam Altman has declared a code red to focus all resources on improving ChatGPT and to stop the loss of users. In this video we break down the drop in traffic, the pause on agents and advertising, and the rumors about a possible GPT 5.2. We also review the new GO subscription, the computing limitations and the importance of voice and personalization compared to what Google is preparing with Project Astra. In addition, we look at Deep Think, Visual Layout, the progress of DeepSeek 3.2, the AI trained on prison phone calls, Microsoft’s VibeVoice model, and the updates from Runway and Kling. We close with the Challenger Report, the labor impact of automation, the MIT safety ranking and Demis Hassabis’s view on the role of LLMs in AGI. 🔗 Useful Links OpenAI code red, The Information: https://www.theinformation.com/articles/openai-ceo-declares-code-red-combat-threats-chatgpt-delays-ads-effort OpenAI reaction to Gemini 3, Mashable: https://mashable.com/article/openai-code-red-reaction-to-google-gemini-3 GPT 5.2 rumors, The Verge: https://www.theverge.com/report/838857/openai-gpt-5-2-release-date-code-red-google-response Viral Gemini Live oil change demo: https://x.com/kimmonismus/status/1995900344224907500 ChatGPT pricing comparison: https://openai.com/business/chatgpt-pricing/ Deep Think for Gemini 3 Pro demo: https://x.com/GeminiApp/status/1996656314983109003 Deep Think explained by Google: https://blog.google/products/gemini/gemini-3/#gemini-3-deep-think DeepSeek V3.2 model: https://huggingface.co/deepseek-ai/DeepSeek-V3.2 DeepSeek pricing: https://api-docs.deepseek.com/quick_start/pricing DeepSeek official web app: https://chat.deepseek.com/ AI trained on prison phone calls, MIT Technology Review: https://www.technologyreview.com/2025/12/01/1128591/an-ai-model-trained-on-prison-phone-calls-is-now-being-used-to-surveil-inmates/ Microsoft VibeVoice Realtime 0.5B: https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B Runway 4.5 video model: https://x.com/runwayml/status/1995493445243461846?t=B5gVVJcgZWGYayqboDXCww&s=08 Kling o1 video examples: https://x.com/Kling_ai/status/1995506929461002590 Kling 2.6 with audio: https://x.com/Kling_ai/status/1996491038937206867 Challenger job report United States: https://www.challengergray.com/blog/challenger-report-71321-job-cuts-on-restructurings-closings-economy/ AI Safety Index by MIT and Max Tegmark: https://futureoflife.org/ai-safety-index-winter-2025/ Demis Hassabis comments on AGI: https://x.com/slow_developer/status/1996975351080034659?t=Ho8TD-uaCOcLAkQ1Ve8mvg&s=08 ________________________________ ⏱ Timestamps: 00:00:00 Intro 00:00:11 OpenAI declares Code Red 00:01:27: GPT-5.2: a potential imminent leap 00:03:20 Deep Think in action: real-world reasoning with Gemini Live 00:05:46 DeepSeek 3.2: the open-source model challenging the giants 00:08:14 An AI model for monitoring inmate communications 00:09:17 VibeVoice: real-time voice generation on-device 00:11:24 Runway 4.5: incremental improvements and benchmark debates 00:12:48 Kling o1: improvements in consistency and visual reasoning 00:13:51 Recent employment shifts driven by AI 00:15:53 How AI labs score on safety performance