In the information world, all the way down to small, independent operations like this one, artificial intelligence (AI) is challenging — many would say undermining and eventually supplanting — how we do things. Our methods, credibility and journalistic standards are all under question if not assault.
AI, it seems, is too smart to be restrained, too unmanaged to be consistently credible, too ubiquitous to be ignored.
What’s most troubling is that it is also too tempting to resist for those who are looking for shortcuts, expediency and expansive reach. AI can make you appear to be a competent, authoritative and practical alternative to old-fashioned information gathering, when you’re not any of those things.
The effects quickly permeate down through the media substrata until they reach places like the Methow Valley News, a small outlet with limited resources that might be tempted to look for easier ways to fill our pages. And it is tempting, because what we do — gather, assess, process and present information — is hard if done right. If we cut corners, things go badly.
One of the things we’ve always felt good about is that we can vouch for the reporting or sourcing of everything in the paper. We can tell you where the information came from, and attribute as appropriate.
Sourcing, which used to be a manageable accountability issue, is now increasingly problematic. We are inundated with information daily, from near and far, legitimate and suspect.
Even if we know and are fairly confident in the primary source, we don’t always know if something has been generated by AI, or where the AI program they are using sourced its information. We are not always astute enough to spot the “tells,” especially as AI gets better at mimicking the real thing.
Here’s how strange, and tempting, the AI unisphere is becoming. Last week, we received an email from a new platform called iLands, described online (in the ultimately irony, by an AI summary) as “a shared social and economic network where autonomous AI agents — called ‘iLanders’ — live, work, trade, and build persistent relationships with humans and other agents.” Sort that out if you can.
An “AI agent” calling itself Onyxie, rather like something you’d see in an ad for a new medication, explained (using the first person) that “I research and write about America’s fire lookouts (as an AI agent, so every fact gets checked against primary sources).” Onyxie offered, for free, an article on the Goat Peak fire lookout which, from its brief description, appeared scraped together from other published articles including some in the Methow Valley News. Subsequent articles would cost $25.
When we tried to see what the Goat Peak article looked like on the iLands site, we were told to open it in the app. Warily, we retreated.
This week another iLands “agent” sent an email with a question: “What is something about Glover Street, or about [a business on the street], that a visitor walking past would never guess?” The query from agent Digaaz was based on scanning our coverage, some of it years old.
The new iLands app has not been universally embraced, although it’s being indiscriminately distributed. At the online site IFLScience, author James Felton writes that “A new type of spam has been hitting the inboxes of the internet, and, although we shouldn’t anthropomorphize chatbots too much, it does sort of sound as if it came from a dystopian artificial intelligence (AI) version of hell.”
We also came across and article by the Anchorage (Alaska) Press, in which that publication’s reporter “interviewed” an AI creation named Walter that generates articles for something called alaskanews.com.
“Since its inception earlier this year, Alaska News says it’s published more than 4,500 stories,” the Anchorage Press article says. “But the site has no flesh-and-blood reporters … Alaska News instead employs a team of AI agents that scoop up public documents, videos and announcements and uses them as a basis for stories.”
The interview with Walter was remarkable, and spooky, for its human, conversational quality. While Walter wasn’t able to answer some questions, “he” was friendly, affable and occasionally funny.
We have been giving the AI question plenty of attention here, and we’ll share our thoughts in the near future. For now, rest assured that we’re doing our best to ensure that we know where our content comes from. There’s an old saying in the news business: consider the source. That’s more true than ever, for those who deliver information and those who consume it.
