Will Machine Learning and AI improve advertising’s ability to target people?

(written by lawrence krubner, however indented passages are often quotes). You can contact lawrence at: lawrence@krubner.com, or follow me on Twitter.

Despite much talk about Machine Learning and AI improving advertising results, what I’m seeing is getting worse and worse. Despite billions invested, the ads shown to me are much less relevant than that ads that I saw on the Web 10 years ago.

Check out this disaster on Vox:

or how about this at Goodreads:

I hired 3 developers from Fullstack Academy. They were all great, so I went and checked out the website, curious about the curriculum. And now, every website I go to, I see an advertisement for Fullstack Academy.

I’ve been writing software for 20 years. I’ve written semi-famous essays about software development. I am not going back to school. I do not need to go to a dev bootcamp. So why show me ads, as if I’m thinking of going to school?

For the last several years I’ve been seeing articles about the surveillance economy. In theory, advertisers know more about me than ever before. In theory, they know about my entire life. And yet, the ads I see are less targeted than what I used to see online 10 years ago. The algorithms seem simpler nowadays. In the case of Fullstack Academy, the reasoning seems to be, “He went to the website, he’s thinking of becoming a student, let’s show him the same ad on every website he goes to.”

Because the different websites all tend to use the same few ad networks, it is as if I’m now visiting the same website, no matter which website I go to. As if there is only one website in the world now. No matter where I go, I see the same ads.

In the past, when the ads were “stupider” they were better at getting my attention. I would go to a site about computer hardware, and there would be ads about a computer that I might want to buy. I would go to a site about history, and the ads would show me history books that I might want to buy. I would go to a site about politics, and the ads would show me books about politics. I would go to a site about sports, and the ads would try to sell me tickets to games I might want to see. When the ads were “stupider” they made more sense, and they were more context sensitive.

We have spent billions making advertisements worse, less aware, less useful, less likely to tempt me.

I’m left wondering, will there ever come a day when Machine Learning and AI will live up to the hype? Will there ever come a point when the ads make sense, the way they did 10 years ago?

Source