Exploring Text Annotation Techniques using OpenAI and WeWeb

In this meeting, the State Changers participants discussed issues about tech sanitation and inputs using OpenAI. One participant shared how they’re using OpenAI for text extraction and running tests for CSS and ScriptTag. There was a focus on the functionality of OpenAI, which the participant described as a minimum viable product (MVP), acknowledging it may not be the best tool for the task but a good starting point.

Discussions included utilizing the Lambda function for concatenation of entity text, and some challenges around code errors when formatting on the backend. There was an emphasis on understanding the limitations of these approaches and a focus on achieving extensibility throughout the function stack. The meeting involved suggestions on how to progress and validate next steps, with emphasis on learning from each version and not trying to avoid rebuilding. There was also a discussion on ways to handle and annotate decorated texts, with a suggestion to handle the formatting in the frontend versus PHP backend for easy manipulation of HTML. The conversation also moved to a discussion around WeWeb and its community. One participant explained the differing opinions within that community about the best way to solve problems—with some advocating for hacky, quick progress, and others advocating for a more thought-out, comprehensive approach. In the latter part of the meeting, the problem around the PM method was discussed, regarding issues with integer vs. text handling in JavaScript, specifically card number handling. A solution had not yet been found, with further exploration needed. The meeting ended with an intention to continue pushing the boundaries of workflow. Keywords used in the meeting include: OpenAI, CSS, ScriptTag, Lambda, WeWeb.

(Source: Office Hours 6/21 )

State Change Members Can View The Video Here

View This Video Now

Join State Change Risk-Free