What this is: A list of a thousand interesting people, with a very rough ranking to help you see very interesting people first. What this isn't: Comprehensive, accurate, or a performant website.
A couple of days ago I saw the Times 100 Most Influential list for 2025, and wanted a bit more: what about a list of influential period across all of history? I searched for some, but there weren't any other lists that met my bar (or didn't seem to have broken websites). I want to learn more about very interesting people in history, so this was a shame! To help me in this pursuit, I decided to make my own version of this list through some combination of wikipedia's existing data and some LLMs. The final ordering I've ended up is coorelated with how I would order people, but does indeed contain many ordering crimes. I hope that at the very least you can find some interesting person here you didn't know about previously.
To generate this list, I first got a big list of people from Wikidata (which is an amazing dataset) who were born after 2000 BCE. I took the top thousand people (out of ~7.1 million) who have the most links to their Wikipedia page from other Wikipedia pages and then used some LLMs to rate each person on multiple different criteria that I'm interested in. These criteria represent an individuals contributions to science, politics, and culture. Of course, defining and measuring influence is tricky and I've only scratched the surface of capturing the scope of people's impact correctly. I've included my own scoring function from these criteria below (and also use the number of Wikipedia backlinks as a metric), but you can also use your own functionto modify the rankings.
If there is someone missing here that you think is a very interesting and influential person in history, I would love to hear! Shoot me an email. As an incentive, I might just add them to this list :)
I asked Claude 3.5 Haiku, GPT-5-nano, GPT-5-mini to rank people on four criteria from 0-5: i) political influence ii) scientific influence iii) cultural influence. For political influence specifically, I used two different sysprompts to reflect a ranking from someone who believes in a Great Man theory of history (to measure more direct impacts), and someone who believes in a Structuralist theory of history (to try to measure counterfactual impacts more strongly). You can find my prompts in my codebase here. I take the average of the three LLMs to get a final score for each metric (the text when you hover over the score is GPT-5-nano's reasoning).
If I wanted to spend more time on this I would probably try to parse out counterfactual impacts of individuals. This is tricky and historians debate this all the time, but it would be sick to get a great version of this. Also, I would try to make the LLMs better at modeling the scope and scale of a person's contribution (for instance, deeply changing the way one hundred thousand people live seems more important than making a movie that millions of people watch and forget about, but this is slightly more tricky to model).
My current formula is slightly weighted towards my interests: I put a premium on the tails of individuals, and slightly emphasize political and cultural influence. Additionally, for political influence, I give equal weight to the Structuralist and the Great Man views of history.
Yeah. I could put it online if there was interest. You could regenerate it using my code, though it's slighlty annoying. Wikidata is sick though!