PewDiePie Trained an LLM That Outperforms Major Models

Press Space for next Tweet

PewDiePie trained his own LLM and it outperforms DeepSeek 2.5, gpt-4o, and Llama 4, and he burnt a GPU and almost his house in the process. at this point he’s not competing with AI influencers, but with AI researchers. he might single-handedly pop the AI bubble.

Topics

artificial intelligence machine learning large language models ai research content creators social media technology

Read the stories that matter.The stories and ideas that actually matter.

Save hours a day in 5 minutesTurn hours of scrolling into a five minute read.