3 ms·

DeepSeek/whoever training on OpenAI outputs is ... bad. OpenAI training on every content creator's outputs is ... good.

• visarga 2y ago
  You say that, but the reality is that all open models rely heavily on synthetic data generated with ChatGPT. They don't like it, but it happens anyway. You can't really protect a public model from having its outputs exfiltrated. This started in 2023 when LLaMA 1 was released, and has been going strong ever since. How strong? there are 330K datasets on HuggingFace, many of them generated from OpenAI.
  • insane_dreamer 2y ago
    Did you miss the sarcasm?
  • Xelynega 2y ago
    And where did OpenAI get the data to generate those datasets?