Can AI help local news figure itself out?
Talking to Paul Cheung of the Knight Foundation about the “present and the potential” of AI for journalism.
The coming wave of artificial intelligence is often thought about in a threatening way, as if truckloads of robots will come to take jobs from grocery store clerks to journalists.
But journalism, with many outlets in dire straits before those real or imagined waves of machine replacements arrive, is also a place where AI can be used to turn things around, from pulling information out of documents to tools like Deepnews.
Outlets such as the New York Times already have their own machine learning teams, but some such as Paul Cheung, Director for Journalism and Technology Innovation at the Knight Foundation, are looking to make sure that advances can help those who are struggling in addition to those doing well.
“You know, from an editorial standpoint, no newsroom is getting tons and tons of investment and resources. It continues to be a challenge, and pressures,” Cheung, whose organization is giving out $3 million in funds around the issue, told me this week.
“I think if they are able to leverage technology properly, they will be able to do more with less.”
Accompanying Knight’s announcement is a report on the “present and potential” of AI for journalism. Among its findings is that of the AI projects being put to use in newsrooms, almost half were focused on helping reporters report their stories.
This makes sense considering that the projects surveyed were mainly from bigger organizations, though shows that there is a gap of using AI to address the most pressing problem for many outlets, which is not that they aren’t doing good journalism but that the business model around their excellent content is no longer working. The issue has become part of mainstream discussion as local papers and journalists worry about owners that don’t want to invest in their work but squeeze out any possible remaining profits while selling off pieces.
Initiatives focused around local news can help find solutions including tech to change the industry, though Cheung also warns of the “copycat syndrome” that can make editors glom onto one tactic as a cure for all their ailments.
“Just because the AP uses NLP to do an automated story, it doesn’t mean you should do that. Just because New York Times is using AI for commenting, it doesn’t mean that it would work for your community,” he said.
I am generally in agreement with Cheung, though for me, using AI well in newsrooms, and technology more generally, boils down to having and using the right data, and this is something that applies to all newsrooms big or small. Tools like Deepnews quality scores and other AI metrics could be useful in conjunction with the mountains of data that local newspapers already have about what people are reading, what leads to subscriptions, etc.
Especially as the internet approaches it bright, beautiful, cookieless future online, there will be a refiguring out how to (and who, besides the platforms, can) make money from writing online. It’s a place where local news can also reinvigorate itself to become a bigger part of the conversation by offering subscriptions to people who want something different than news on Facebook or through search.
Personalization was one of the areas where Cheung pointed out how AI could help readers have a better experience with local news, with a newspaper site perhaps acting more like a Netflix queue or Spotify recommendations in that they can provide people stories based on more complex factors than just category, such as serving someone another business story if they are reading a business story.
Another trend is discussions of data and the ethics around it, which gets into questions about who has your data, what they are using it for, and who you are comfortable having your data (perhaps you’d prefer a local newspaper rather than a bigger company or a data broker).
“One of the hopes is that people will be a lot more educated, the general public will be a lot more educated about this technology,” Cheung said.
“And I think journalism to be sort of a leader in shaping that conversation and the ethical use of it.”
