Uploaded April 2023 | Updated September 2026, 2 weeks ago
Sure, you can use the OpenAPI API to send davinci some text and ask it to extract the most logical keywords representing the topics it should be indexed on for search. However, you'd have to pay the API cost. You could do it manually copy/pasting into ChatGPT. Or you could try to get Copilot to do it somehow or other if you already pay for Github Copilot.
Or you could just do it with Python's scikit-learn package, or with various other approaches that don't actually require a large language model (LLM) chatbot like OpenAI GPT-3 or GPT-4. It's just statistics and regressions and stuff like that which we've been doing awhile in the Python world. As with so many things in Python, it's not a question of whether you can, but in which way you wish.
This is my first attempt. While less than immediately actionable results, it sure is a pretty hierarchical tree it drew. Wish I could read that.
Sure, you can use the OpenAPI API to send davinci some text and ask it to extract the most logical keywords representing the topics it should be indexed on for search. However, you'd have to pay the API cost. You could do it manually copy/pasting into ChatGPT. Or you could try to get Copilot to do it somehow or other if you already pay for Github Copilot.
Or you could just do it with Python's scikit-learn package, or with various other approaches that don't actually require a large language model (LLM) chatbot like OpenAI GPT-3 or GPT-4. It's just statistics and regressions and stuff like that which we've been doing awhile in the Python world. As with so many things in Python, it's not a question of whether you can, but in which way you wish.
This is my first attempt. While less than immediately actionable results, it sure is a pretty hierarchical tree it drew. Wish I could read that.
