I have been writing about and using Google for decades, so I get it — it’s super-easy to use Google for quick one-off searches like finding official web sites, getting a name spelled correctly, getting a quick topic overview, etc. But with Google’s insistence on using water and electricity on AI to summarize the information that can be found right underneath in the first search result’s Wikipedia page, that habit feels wasteful. To me it feels irresponsible.

When I developed MiniGladys ( https://megagladys.com/mg/ ), it was with the idea of making a tool that extracted and presented Wikipedia data in an easy-to-use way. As I have used it over the last couple of months I’ve found it’s replaced about 90% of my Google quick reference searches. What’s the official web site of this organization? What’s the Twitter account of that celebrity? What’s a quick overview of this topic? Has that topic been in the news lately? How can I build a set of web searches for the other topic?

The best part is that MiniGladys doesn’t use any AI; it’s all Wikipedia API and a bit of data analysis. I’d like to tell you a little bit about it and invite you to try it; MiniGladys is free to use and free of ads. Here are three use cases for MiniGladys quick search.

Quick Search I: Finding an Official Web Site

If I need to find an organization’s Web site, I first check MiniGladys. This morning I needed the site of the Competitive Enterprise Institute so I plugged it into MiniGladys. (The search form is autocomplete so if you’re not entirely sure of the spelling, just start typing and you’ll get suggestions.) Wikipedia had a page on the organization. The front page of MiniGladys shows official information about the topic, like official web site and social media accounts and authoritative external links from places like the Library of Congress. I click the Official Website link and away I go, no Google necessary.

Quick Search II: Getting some backstory

Sometimes you’ll see an unfamiliar name in the news, or a name you know vaguely but have no context for. If you do a Google search for the name you’ll often get recent information, but it gets a little tougher to find more historical news. The Gossip Machine portion of MiniGladys uses Wikipedia page view data to find dates where the topic you’re interested in was a subject of unusual public interest, and turns those dates into Google searches. This portion of MiniGladys doesn’t always work, especially if the Wikipedia topic in which you’re interested gets relatively few page views. But it works great for celebrities, nationally-recognized politicians, topics in the news, etc.

Take Mike Flood for example. Mr. Flood is a congressman from Nebraska who had a very contentious town hall recently. I had heard his name but I didn’t know much about him, so I opened up MiniGladys’ Gossip Machine tab and searched for Mike Flood over the last 5 years. All you have to do is click a button.

MiniGladys performs its analysis and shows you some statistics on the average page views for your topic (Mike Flood’s are comparatively low because he didn’t become a national figure until 2022.) Dates of particular public interest are shown with bars and z-score indicators. A search link with each lets you search Google news for that topic on that date. In Mike Flood’s case, I sorted his results by highest spikes and learned something very interesting.

The highest spike of public interest was August 5, which had to do with the raucous town hall. But the second-highest spike was March 19. Can you guess what that was? I couldn’t so I clicked on that result’s “Google News search for this date” and….

The March 19 public interest spike was from an EARLIER raucous town hall! That gives me a bit more context and more information with which to explore Mr. Flood’s political endeavors further if I wish. It also teaches me something I couldn’t have learned using Google News. Here are those two big spikes in the results.

The most recent town hall, which presumably caused the spike of public interest, generated 18.857 views of Mike Flood’s Wikipedia page on that day which was a comparative z-score of 37.80. (That is very very high.) In comparison, the incident around March 19 generated (presumably) 6,703 page views and a z-score of 13.28. That tells me there was apparently a *lot* more public interest in the recent Mike Flood town hall than there was in the one in March. I wonder why? It’s not a data point I can use immediately, but it’s a fact I would keep in my back pocket if I chose to do more research on Mike Flood — I might keep my news searches more recent or double check early August’s events to see if there was anything else that brought his name into the news.

Quick Search III: Building Contextual Topic Queries

Sometimes I run across topics that I want to take a few minutes on but for which I don’t want to do a bunch of searching. In that case I use the MiniGladys Related Topics tab to the see the idea from different angles. Resources are also there for Google and Google News searching.

Recently I was exploring the idea of nature-based solutions a bit and I wanted to get an idea of related topics so I could build better queries. I looked up the page on MiniGladys and hit the Related Topics tab. The Related Topics tab analyzes the other Wikipedia pages which link to your topic and gives you a list with attendant links for Google and Google News searches.

Here’s what the nature-based solutions related topics list looks like:

The related topics are listed by the number of mentions they have and you can filter for a minimum number of mentions. (General topics like beer will have tons of related topics unless you set the filter to something like 10.) Nature-based solutions had related topics like Climate change adaption, Sponge Cities, Business Action on Climate Change, and Land Recycling. Each result is accompanied by link to search Google and Google News for your topic along with the related topics.

Conclusion

MiniGladys uses no AI. Instead, it uses the human-developed structure of Wikipedia data and a roadmap of public interest using the persistent metadata of time. These searches use context methods that AI search often lacks. Further, Wikipedia’s information structure has been developed over decades. We can easily take advantage of it to build better and more complex queries without using AI. Why don’t we?

Before I end this article I’ll note that there’s one overarching question for this article that I never addressed: what happens when the topic I’m interested in doesn’t have a Wikipedia page?

That’s when you need to do some sideways searching. We’ll talk about that tomorrow.