A week ago, I purchased a subscription to a site containing academic resources and wanted to share it with a friend, as well as save it for myself so I could cancel the subscription. However, the resources on this site were scattered across many pages. At first, I manually went to each page and pasted its contents into a document file, but this was tedious. As a solution, I wrote a python script to open a browser using Selenium, extract the useful text, and add it into a .csv file.
Two days later, one of my school teachers gave an assignment that involved following a series of simple instructions (like inputting numbers into websites). I remembered the script I had written two days ago, but that wouldn’t work because each instruction used a different website with different things to do. I would need something that could interpret directions and intelligently choose what to do in a given webpage: an LLM.
This article is part of AI Automated Browser