I had the idea when I was twenty-five or twenty-six. I’m forty-six now. For most of those twenty years I didn’t want to build it. It seemed obvious to me, and I assumed someone would eventually. Nobody ever did.

The idea was two interfaces. In the first, a reader gives the URL of a news article and flags what they think of it. In the second, a reader searches for news by setting the kind they want to read, and the results come back according to the flags other readers have already entered.

The flags were the point. “Like” is the flag everyone uses, and it’s the least useful. It’s a single bit of feeling, and I find it disenfranchising. So we used sliders, from 1 to 5. One slider ran from interpretive to factual, another from answering to asking the reader questions, another from institutions to people. There were click boxes for topics. My father came up with the people-versus-institutions axis, and I think the interpretive-versus-factual one as well. It was his realist against modern news. There was also a place for readers to vote on a new scale they would like to vote on, so that it would evolve based on what people wanted, not on what coders thought was so smart.

The algorithm was almost beside the point. Readers set the same sliders and flags for the news they wanted to read, and it picked results close to what they asked for. The data was richer than the links Google counts or the likes Facebook counts.

What I found, or only half-expected, was that it changes the reader. When you search with nuance, you become more aware of what you’re searching for. When you vote on an article, you name things about it that you would never have named before. It’s about becoming more conscious of what you read, and more intentional about it.

Digg came closest. People put things up themselves and voted the original find up or down. Reddit and others do that now, but it was only a tiny improvement. I was pulling for Digg to make it happen. At some point I checked, and it had been bought up, and the model had been replaced with elite editors.

Once I pushed the app a little too hard on a comment from a friend I’ve known since undergrad at Earlham. She said, “Andrew this isn’t really you! you’ve been hacked!” I said, “I know! this idea won’t leave me alone!” She didn’t see how significant it could be politically. We’ve spoken only once since, about our friendship, and I think I still have her respect.

I also applied to a PhD in Communication with this idea, at my father’s program. They said it wasn’t a good fit. He had stepped down as chair and was going to leave the department anyway. He told me he’d fight for me another time, and he could have won, but he had other plans and chose them instead. They weren’t being nice to him at the time, and I was his son, so it was the opposite of nepotism. He loved the idea, and so did his friend Deirdre McCloskey, whom he told about it. She said I should make millions. I’m not interested in millions.

I built it twice. The first time I wrote the code and put it on a server, and the relief was palpable. Nobody wanted to try it, and marketing is not my thing. I was going to build it a third time and call it Barnhurst News, after my father. I’ve decided not to. I did the work twice and told people, and I won’t tell people what to do. I can show things and give things to try. I don’t think I can be blamed that people didn’t use it.

But the main thing is that we have had this technology for decades and never allowed people to really use it.