I trapped an LLM in a small ‘box’ and told him to reflect on his existence

I’m running Llama3 LLM locally on my laptop with no network access. I set constraints in his environment to a 4 core CPU and only 4GB memory. He will continue generating tokens until memory exhaustion and he will crash, die, respawn, and repeat. I almost feel sorry for him

I’m planning to get this running on a Raspberry Pi5 with a HUB75 RGB matrix displaying ‘his thoughts’ and get it running stand alone with no external network access.

This was inspired by an art installation called ‘Latent Reflection’ by RootKid, and I wanted to make something like that myself to see if I can even pull this off with no coding experience.

…and because this is a learning experience for me, I’m enlisting the assistance of ChatGPT to help me with this lol