AI subscriptions add up faster than you notice. I'm paying $236/month across seven AI tools — and every company I'm giving money to says they're losing money on me. The math only works because of VC subsidies that won't last. When the bill catches up, you'll wish you'd built your own stack. This playlist tracks what AI actually costs, why it's about to cost more, and what happens when you stop renting and start owning your compute. Topics covered: • What you're really paying every month (the subscription audit) • Why "$20/mo" turns into $200/mo (the enshittification cycle) • Self-hosting the AI tools you actually use • Replacing SaaS with a homelab — what works, what doesn't • When AI is genuinely cheaper than humans, and when it isn't If you've ever stared at a credit card statement wondering when "just one more subscription" became $100+/month, this playlist is for you.
Curated by: Codacus (15 videos)
Run a 35B parameter AI model on just 6GB VRAM using llama.cpp and Qwen 3.6. This setup shouldn’t work—but with the right optimizations, it reaches good enough tps on a GTX 1060. In this video, I break down how to run large language models locally on low VRAM GPUs using MoE offloading, memory tuning, and a few critical flags that dramatically improve performance. What you’ll learn: • How to run 35B LLMs on 6GB VRAM • llama.cpp optimization techniques • MoE (Mixture of Experts) offloading explained • Fixing slow token generation (3 tok/s → 17 tok/s) • Using --no-mmap and --mlock for performance and stability • TurboQuant for increasing context length • What doesn’t work (and why) Hardware used: • NVIDIA GTX 1060 (6GB VRAM) • Intel i3-8100 • 24GB RAM Tech stack: Proxmox → LXC → Docker → llama.cpp (adapt based on your setup) Useful resources: • Qwen 3.6 35B-A3B model: https://huggingface.co/Qwen/Qwen3.6-35B-A3B • TurboQuant paper: https://arxiv.org/abs/... • llama.cpp TurboQuant fork: https://github.com/TheTom/llama-cpp-turboquant If you're interested in running AI locally, optimizing LLM performance, or pushing old hardware to its limits, subscribe for more experiments. Chapters: 00:00 This shouldn’t work 00:27 Setup 01:46 Why it’s slow by default 02:52 MoE breakthrough 04:33 Fixing memory bottlenecks 05:32 Hitting 17 tok/s 06:40 4× context trick 09:23 Stability fix 11:04 What failed 13:32 The 5 flags #LocalAI #LLM #llamacpp #Qwen #AIonGPU #LowVRAM