Last week I was working on a data project. I needed to get information from a website and manually enter it into a spreadsheet. In the past, I’ve used Upwork to find data entry resources to help with this sort of thing, and I did that again this time. But as they worked (I have them do tests before I give them the full project), I soon realized that at the rate they were working, it would take someone weeks to build the spreadsheet and cost several times more than I wanted to pay.

I wanted to be frugal and get this task done faster, so I reached out to a developer friend for suggestions. He showed me how to extract the data using the website’s API instead of the user interface (i.e., webpage). All I had to do was figure out how to parse and store the data. Normally, I’d have a developer do this for me, too, but I turned to AI instead. I cracked open Google Gemini and explained the project. Within minutes, it had written a Google Apps script that I could run. It would take the data returned by the API, parse it, and store the data in Google Sheets. The script required one manual step, too, but I was happy to do that. It was very efficient and way less prone to error as compared to someone manually doing it over a few weeks. So, I spent the weekend using the script, Gemini, and the website’s API to build a large data set.

In the end, a project that was projected to take weeks, cost a significant amount, and require me to manage another resource was instead completed by me (with help from Gemini) in several hours over the weekend.

Once people realize they don’t need developers to write code and can build things on their own with AI tools, we’ll see an explosion of people building tools to solve their own problems (and do their jobs more efficiently).

‍