I killed an AI idea today

Little things in life… but I killed an AI idea today at work, and I am pleased about that. 😀

A colleague reached out with a requirement he got from someone else. They wanted to create an AI agent that will scrape the Internet for certain types of news and put it into a SharePoint site.

I looked at the problem differently and said this doesn’t need AI. In fact, this might be an example of how we shove AI into everything nowadays, often not even choosing to look at alternatives. In this case, what do they mean by “news” to begin with? Typically that would be search results in Google, so I said why not just setup an alert in Google for the keywords we are interested in. This has existed for ages, and the neat thing is it can send an email or create an RSS feed of the results (ah good old RSS feeds!).

All one really needs to do then is create one or more alerts like these, and then work off that. In the past one might have created a Python script or something, but now use Power Automate to loop over the RSS feed results, extract the title, the URL, and so forth… and dump them into a SharePoint list.

If you must introduce AI at this point, I suppose you could pull the content and get AI to summarize it or do some sentiment analysis, but does it really matter. Why waste money and contribute to global warming for no reason.

ps. I have nothing against AI or using it when needed, but I am against trying to make AI the one and only solution for every problem we have.