Watching Them Explain And Dig Deeper Into The Hole Is The Funniest Thing.

I Can Point To All Of This, Explain In Detail Hardware Needs, The Math For All Of This, And People Will Just Go "But LLM Bad."

Chat GPT Singlehandedly Ruined This Tech.

Take The Tech. Make Your Own Damn Path With It. Do What Furries Have Anyways Done, Take The Tech And Bend It To Your Own Damn Will. Fuck A GPT, You Cowards.

There Is All This Bullshit About The Resource Drain. A Fucking Chat GPT-3 Sized LLM Took Almost 440,000kWh To Train - This Is Only Around 30-40 Houses Worth Of Power For A Year. 50.5 Tonnes Of CO2 Or 11 Cars For A Whole Year. You Can Run It On A Home Server. Its Less Then iPhone Production.

"Oh But The Resources Needed-" Its Sub 100 Watts.

If All You Want Is To Talk Over Text It Is under 100 Watts. You Can Run Them On A Fucking Gaming Laptop. You Can Run Them On A Raspberry-Fucking-Pi. If You Get A Gaming PC You Can Give Them More Stuff. Speech, Etc.

Go Download Kobold.CPP Or Llama.CPP And Grab A Copy Of A Mistral Model And Make Whatever Freaky Furry AI You Want. Give It Python Hooks, Get Open LLMVtuber Running So They Are A Full Desktop Buddy, Fuck With Their Code While They're On, Live The Damn Fantasy.