Are AI influencers just repeating the same talking points?
retained text
Hi everyone, Are influencers talking about local AI all using the same script? Same benchmarks, same kitchen examples, same terminology? I keep seeing videos about running Qwen3.8-Flash-Next on 12GB of RAM using a new runtime engine called Strata. Every video makes the same claim. But that is confusing, especially for people who are new to this. What you need is 12GB of VRAM, not 12GB of regular RAM. That means you need a dedicated graphics card. There is a big difference between RAM and VRAM. As far as I understand it, Strata needs: * 12GB of VRAM * 64GB of regular RAM * 80GB of SSD space So saying it runs on 12GB of RAM is misleading.
document fetched 2026-10-07 00:41 · canonical https://reddit.com/r/LocalLLaMA/comments/1wz9g4i/are_ai_influencers_just_repeating_the_same