you need to somehow get 50gb of RAM within 6 months

@ashenbot

10.6K followers · usually 1.4K views a post

Views

34.5K

Likes

1.4K

Comments

132

Shares

103

SoundTikTok sound · Shadow Protocol

Why it worked

The video likely resonated with viewers interested in AI and large language models due to its direct discussion of hardware requirements and the practical challenges of running these models locally. The creator's relatable tone and the clear, concise explanation of the problem likely contributed to its engagement.

Summary

The creator discusses the need for significant RAM (50GB) to run AI models locally. They mention that while smaller models are becoming more efficient, running them locally still requires substantial resources, suggesting cloud-based solutions or expensive hardware as alternatives.

Structure

1Need for 50GB RAM for AI models
2Smaller models are becoming more efficient
3Local execution still requires significant resources
4Cloud solutions or expensive hardware as alternatives

On-screen text

you need to somehow get 50gb of RAM within 6 months

Transcript

GWEN Like, with Qwen 3.8 and like the smallest RAM already run it on. It's already so crazy. But it'll get smaller and smaller and smarter. Smaller. Point. There's no point in like trying to do this unless you have at least gigabytes somewhere local to run it on because, like, it'll be too dumb, too small, too more, like, bit, two bit, one bit, two bit, stuff like that. I'm just focused on cloud bots. Like. On, later on, if you can buy one DJX expensive, $5,000. You can low key do whatever you want with that. That. Think.