… And unlike ChatGPT, both the AI models and the knowledge base they can harness remain on my home network, so I don’t have to worry about cloud-based clankers gaining access to personal documents. …
… No longer do you have to wait, see the results, decide you're not happy with them, and then wait all over again. Instead, you can have live results that you sketch and adjust as you need. You also get access to inpainting and outpainting. …
… I stuck with gpt-oss because it’s an open-weight model from OpenAI, so you’re getting the same training quality, just without the cloud and internet dependency. …
… A 7 billion bytes of training data B 7 billion parameters C A model trained for 7 billion steps D 7 billion tokens in its context window Correct! …
… So I have RAG on my document, the model’s training knowledge, and live web access in the same session - without having to tool-hop anywhere. …
… Projects organize context with persistent file uploads Your documents stay attached to every conversation ChatGPT trains you to think in conversations . You open a new chat, get answers, then usually start fresh the next time because you're trained to follow that routine. …
… Every model made the same mistakes Training data biases were too strong There's a pattern that runs through all five models, regardless of size or architecture, and it's down to training data. …
… A 7 billion bytes of training data B 7 billion parameters C A model trained for 7 billion steps D 7 billion tokens in its context window Correct! …
… I wanted the power of modern AI to summarize and brainstorm, but I wasn't willing to trade my privacy or rely on cloud services that require constant internet access and data sharing. …
… A 7 billion bytes of training data B 7 billion parameters C A model trained for 7 billion steps D 7 billion tokens in its context window Correct! …