{"id":1072,"date":"2022-12-07T11:36:52","date_gmt":"2022-12-07T11:36:52","guid":{"rendered":"https:\/\/www.denizcemonduygu.com\/philo\/?p=1072"},"modified":"2022-12-07T11:36:52","modified_gmt":"2022-12-07T11:36:52","slug":"getting-help-from-ai-to-tag-sentences","status":"publish","type":"post","link":"https:\/\/www.denizcemonduygu.com\/philo\/getting-help-from-ai-to-tag-sentences\/","title":{"rendered":"Getting Help From AI to Tag Sentences"},"content":{"rendered":"

I\u2019ve recently <\/b>announced<\/b><\/a> a new feature I call \u201ctags\u201d <\/b>\u2013 keywords, shown on the left of the sentences, including names of specific arguments, theories, -isms, etc. which don\u2019t appear in the sentence itself. Tagging, I had said, was an ongoing manual process I\u2019d been carrying out for some time. I had tagged 525 sentences myself <\/b>by going through the total of 1603 using the wet, mushy neural networks in my brain.<\/p>\n

Even though I have intentionally avoided using any kind of automatization for my editorial work in this project since the beginning, I got curious about the kind of help I could get from AI applications for the relatively easy task of tagging sentences. So I plugged <\/b>OpenAI<\/b><\/a>\u2019s <\/b>Generative Pre-trained Transformer 3<\/b><\/a> (GPT-3) into my database.<\/b><\/p>\n

GPT-3 is \u201can autoregressive language model<\/a> with 175 billion parameters, 10\u00d7 more than any previous non-sparse language model\u201d. It is described<\/a> as \u201cone of the most interesting and important AI systems ever produced\u201d by David Chalmers, currently the youngest (56) living philosopher listed in my project. GPT-3 has an <\/b>API<\/b><\/a> which returns a completion for any text prompt you give, attempting to match the pattern. The prompt I designed for the task went like this:<\/b><\/p>\n

Sentence: There are two kinds of substance: mind (which I am as a conscious being; \u2018res cogitans\u2019) and matter (\u2018res extensa\u2019).<\/h5>\n
Category: Dualism<\/h5>\n
Sentence: There are no innate certainties to be discovered in our minds, all knowledge comes from experience.<\/h5>\n
Category: Empiricism<\/h5>\n
Sentence: The universe consists of matter in motion, and nothing else exists; it is a vast machine.<\/h5>\n
Category: Materialism<\/h5>\n
Sentence: \u2026<\/h5>\n
Category<\/h5>\n

Thanks to this prompt and a few parameter settings, GPT-3 was able to \u201cunderstand\u201d what I was asking from it and to work just as I intended it to: it read the prompt and produced completions to it 1603 times, each time with a different sentence<\/b> from my collection substituting the \u201c…\u201d in the prompt (thanks to a function I created in Google Sheets Apps Script based on Richman\u2019s<\/a>), thus offering 1603 suggestions for categories. But how successful was it, from a human point of view?<\/b><\/p>\n