AI isn’t going to kill journalism. It’s going to kill bad journalism.

One of the laments I’ve seen recently is that the rise of AI agents is going to kill off publication readership. “If you can get the answer to their query from a Google AI recap or GPT then you’re not going to go to my website to find the answer.” I’ll admit that I was pretty much in agreement up until recently. But in the search for something completely unrelated I had a realization - it’s only the filler that’s not going to be read.

Think about it this way.

When the internet got started the biggest points of friction in terms of content were creation, then discovery. It was hard to make a website early on, and discovery before search engines meant you had to have the exact URL or IP address to access the content. Then the first online revolution happened with the rise of search engines that scrolled, catalogued and presented the contents of the internet to a user's query. Soon after, the second revolution happened and self publishing tools came onto the scene and anyone with an internet connection could share their ideas with the world. It was the great democratization of knowledge we were promised.

Pretty soon after this SEO became a thing and very quickly the practice escalated into a kind of online warfare. People began to try to game the system and get views to their website by manipulating or creating content that was designed to lure people via search engines. This is when every company on earth started a blog on their website and began generating their hot take on the news of the day. To try to get you to their website, despite the lack of relevance to their content or expertise. So search engines would routinely change their algorithms to try to beat the SEO companies and so the cycle would go.

The SEO wars eventually resulted in the current glut of content that exists on the web now. Quite honestly that excess of content is why search results have gotten markedly worse, especially for those that don’t include boolean modifiers in their searches.

Throw in a handful of programmatic ads and Taboola-style ad tech platforms and you had some truly horrible experiences trying to find answers online.

Enter the promise of AI.

AI will give you an answer for your question, without having to dodge past countless popup ads, “download our whitepaper” CTAs, and “read more” prompts. And for questions with an answer it should be a much better experience for the user.

Will this kill “publications” that primarily served as republishing vehicles for bulk ad units? I hope so. Will it kill journalism and reporting? It absolutely shouldn’t. What it has the potential to do is open up the field by pruning the republishers. It opens the door for quality publishers (and by quality I mean those that do reporting and have a distinct point of view) to reach their audience again. And those publishers that take the time to focus on the experience for those users will see engagement rise, and hopefully with it revenue. My hope is that quality publishers will still resist the use of AI in their reporting. AI, no matter how well prompted, has a tendency to strip the point of view from an author and create a sameness. Great journalism brings the conviction of a lived experience that AI can’t recreate. Even the best of it feels neutral at best.

So should publishers fear AI or embrace it? If publishers focus on clarity, quality, and voice, AI won’t be the end of journalism, it might just be its savior.