Given that AI models are trained on Reddit data, isn’t it surprising if someone hasn’t already been shamed or called out as “shittymorph” by now?

Exploring AI’s Familiarity with Reddit Culture: Have Models Encountered ‘Shittymorph’?

Recent discussions within the AI community have raised an intriguing question: Given that many artificial intelligence models are trained using Reddit data, is it possible that some models have been exposed to, or even influenced by, niche Reddit communities like the ‘Shittymorph’ meme?

This curiosity prompted a practical experiment. I decided to test whether modern AI assistants, such as Google’s Gemini, are aware of specific Reddit slang or meme styles. By instructing Gemini to emulate a ‘Shittymorph’ style comment, I observed the model’s response. Impressively, the AI delivered a reply that encapsulated the distinct tone and quirks associated with that Reddit subculture.

This encounter hints at a broader question: How deeply do these AI models comprehend or internalize the diverse, often underground, facets of Reddit’s community-driven content? Exploring more obscure Reddit lore might serve as a revealing probe into the boundaries of what these models have learned.

Are there particular Reddit memes or subcultures you’d like to see tested? Understanding the extent of AI’s cultural knowledge could offer valuable insights into both the strengths and limitations of current language models.