As local newsrooms go, Richland Source is much like any other. They cover several counties in their region – North Central Ohio – and employ people with varying years of experience. They focus on solutions-driven reporting and community involvement. They have brought home “a few” awards.

Their CEO, Jay Allred, leads the agency, Source Brand Solutions, which operates Richland Source and other properties. He also leads the automated reporting and technology company, Lede AI.

Lede AI was built in the newsroom and has been around for about a decade. “The assumption most people make is that we (Richland Source) use AI everywhere,” Allred said. “But the reality is my newsroom isn’t much different than anybody else’s.”

There is no mandate to use or experiment with AI. Allred says there are also employees who are eager to experiment sitting next to those who want nothing to do with it.

He also didn’t want to create an internal team that created a bunch of additional rules or policies for everyone else.

“We already have rules about being good journalists” he said, and he wanted to encourage applying those to the use of AI.

They decided to create a small “Strike Team” that included people who are semi- to very interested in generative AI. They also created a company Slack channel that anyone can participate in.

Here’s how it works: The team selects a problem to solve or idea to work on for a month. At the end of the test period they report the results using a framework that answers five questions:

  • The idea, or problem we tried to solve
  • The tool we used
  • What worked and didn’t work
  • What we recommend
  • What we’re going to do next

“It’s simple and allows for a fast cadence of experimentation and reporting back,” Allred said. “This structure also allows the focus to remain on what a person learned, not whether you, or the experiment, failed or not.”

All of this, Allred said, is rooted in the company’s values to try, learn, and repeat.

“Learning that ‘AI sucks for this thing that I tried’ is a lot different than, ‘I tried to make this work, but I couldn’t make it work,’” he said.

“Humans are adopting this at different rates and with different levels of interest,” Allred said. “And we’re just trying to create a really safe environment to adopt it.”

Other findings

Allred says it’s really wise not to make assumptions about who is going to be your early adopter based on their age or their gender, or what they’re interested in.

The company’s biggest adopter of AI, who uses it most frequently, is also their oldest journalist. He characterized it as a free assistant available 24/7 that can’t do enough for him. Allred said this editor loves reporting but finds the writing process challenging, even though he’s a very good writer. ChatGPT helped him with his writing process, reducing the pain in the process.

Fears about being replaced by technology are also being addressed because the team is viewing experimentation as a reward and opportunity for advancement.

“From a leadership perspective, this makes it really easy for others to enter without a huge set of expectations around what’s going to happen,” Allred said. “And it helps that we’re not firing anyone or laying anyone off around the use of AI.”

A final bit of advice

Let’s say you have five reporters in a newsroom, and none of them are eager to use or learn about these tools right now, Allred said. Start collecting inspiration from other newsrooms who are experimenting and pull it together for them.

The advice can’t come from the leadership because then it’s seen as top down, Allred said.

“But if it comes from some other newsroom, then it’s more encouraging,” he said. “Then they’re looking at their peers who are doing this work.”