LLMs Can't Reason And They Never Will - Here's Why. скачать в хорошем качестве

LLMs Can't Reason And They Never Will - Here's Why.

AI was supposed to make us faster, smarter, and more creative. It did the opposite. In this video I'm going to explain how LLM models stopped improving in late 2024. I'm going to tell you exactly why, and I'm going to prove it. We'll discuss all the various ways AI can damage you - from bad information and hallucinations all the way up to cognitive atrophy and decreased productivity. When you're done watching this video you may just rethink your ChatGPT subscription, and you're definitely going to rethink your OpenAI and Nvidia investments! TLDR: LLM technology CANNOT reason and thus can never reach AGI no matter how much you scale compute! If you enjoyed watching please subscribe for weekly updates. 00:00 INTRO 02:55 THE PRODUCTIVITY MIRAGE 07:05 THE COGNITIVE LOBOTOMY 10:57 THE MATHEMATICAL CEILING 15:55 CONCLUSION #llm #openai #ai #nvidia SOURCES : METR Study about the Impact of AI on Developer Productivity: https://metr.org/blog/2025-07-10-earl... Mckinsey State of Ai in 2025: https://www.mckinsey.com/capabilities... Uplevel Data Labs study - Can Generative AI Improve Developer Productivity?: https://resources.uplevelteam.com/gen... Apple The Illusion of Thinking: https://ml-site.cdn-apple.com/papers/... MIT project study - Accumulation of Cognitive Debt on brain from using ChatGPT: https://www.brainonllm.com/ Epoch Ai study LLM scaling: https://arxiv.org/pdf/2211.04325 Epoch Ai research Ai computing demand: https://epoch.ai/data-insights/ai-chi... The New York Times - When A.I. chatbots hallucinate: https://www.nytimes.com/2023/05/01/bu...