Is the Article Doomed as Video and AI Rise?

By: Jacob Cohen Donnelly

If you want to make anyone in the media industry break out in a sweat, utter the words “pivot to video.” The reaction—especially if they’ve worked in the industry for a long time—will be visceral, reminiscent of the mid-2010s when numerous media companies laid off writers to chase video content dreams.

And so, reading an article on Nieman Lab about how news outlets are investing more in video might shock some people. It’s nearly 2025. Haven’t we learned our lesson? Maybe this time is different. According to the story:

Short vertical videos, a staple of Instagram, TikTok, and YouTube, are increasingly making their way on to news sites. The format is a hit with audiences — the Reuters Institute for the Study of Journalism (RISJ) reported earlier this year that 66% of respondents in its global survey watch short news videos every week — but more than two-thirds of that viewing still takes place on platforms, not publisher websites, RISJ found. Publishers are trying to change that.

…

News organizations are building vertical video carousels for their homepages and apps, embedding vertical videos in stories, and having their video journalists produce both horizontal and vertical content — as well as asking reporters to present their stories on camera. “Prestige TikToks,” The Atlantic’s Charlie Warzel called them last month.

A few key points in the story jump out. First, this is not a platform play. While the publishers are distributing content there, it’s not because anyone has given them any money. Instead, this is a strategy born out of a desire to serve its on-website audience. Many people are visiting these publishers on their phones and they expect a certain user experience, aka vertical video.

Second, these aren’t entirely new teams being spun up to create the video content. Instead, the reporters who normally write an article are being put in front of the camera. These aren’t video-first folks who are perfect on camera and will need help to get better; however, they’ve got the subject matter expertise about these topics and are able to articulate that information.

Third, there’s a clear evolution in how people consume content. People like video. As RISJ reported (in the above quote), 66% report watching short news videos every week. We’re moving into a new world where people get their information across formats. The publishers that can’t satisfy that are leaving a hole in the market.

Rafat Ali, founder & CEO of Skift, a B2B travel publication, tweeted about this on Monday. He calls it the rise of multi-format media:

Platforms like podcasts and of course TikTok have pioneered a seamless, multi-format approach, where users move effortlessly between video, audio and text without even noticing the transitions. This immersive consumption experience strikes a major blow to traditional formats by offering richer, more engaging ways to tell stories.

While he’s right that it strikes a major blow, it does not have to be a negative outcome. If we look at the examples in Nieman Lab’s story, many of these videos are created with a “yes, and” attitude. The reporter still produces an article and then creates a video. The outcome is two pieces of content where the primary input—great reporting—only has to be done once.

But you can imagine extending even farther. Imagine you’ve got a beat reporter who covers airlines (let’s go with one of Skift’s coverage areas). The big scoops could come with a corresponding video and photos, so all that investigative work generates multiple audience touch points. After five or ten articles/videos about a similar topic, it can be repackaged into a longer-form YouTube video along with a downloadable ebook.

Oftentimes, we forget that there are multiple steps to creating content and the reporting or fact-finding is the first one—and, frankly, the most expensive part. We should do everything in our power to extend the value of that hard work. Let reporters spend more time reporting and then encourage them and give them the tools to create articles and videos and podcasts that immerse the reader in the coverage.

It does beg the question. If the reporting is the primary input, how far can you push it? As Ali continues in his tweet:

Tools like ChatGPT provide answers in highly efficient, information-dense formats, which prioritize user-friendliness and retention. While these outputs can feel monotonous or homogeneous, advancements in technology—and the human element that will persist—are likely to solve this problem over time.

…

this shift toward AI-driven formats is reshaping how information is consumed and valued. The story format, in its traditional sense, is being overtaken by new methods of engagement that prioritize efficiency and density.

Could we see a world where there is no story at all and, instead, the reporter spends their entire time gathering facts and putting them into a publisher’s proprietary AI system? Is the final boss of multi-format media all of us asking questions to AI?

There are three answers to this. The first is a simple yes. I have seen a dramatic drop in my Google usage and find myself spending a ton more time on ChatGPT. It gives me answers much faster than I could ever get from an article or even a video. The ability to port in trusted information into an AI platform to get answers to my questions is absolutely a future that I envision.

Second, while ChatGPT answers questions easily, I find that it doesn’t provide much context, which articles and videos do very nicely. ChatGPT can tell me the answer to something, but it can’t contextualize it as part of a broader concept. Only someone who has been covering something regularly—a reporter—can provide the right context at the right time. It’s why reporters get better with time on the beat; they gather more information and understand what matters over time (and make connections).

Third, we have to remember who is creating the inputs for these AI platforms: reporters. Anytime I ask a reporter if they imagine a future where they’re just interviewing sources and not writing articles (something many might relish), they all say no. Their reason? Writing the story is itself a crucial analytical process. The act of structuring a narrative, choosing what to emphasize, and making connections between different pieces of information helps reporters develop deeper insights and identify new angles to pursue. If you’re just taking interviews and piping them into an AI platform, you’re removing this essential analytical step that often leads to better journalism.

How we consume media is evolving. Publishers need to be intentional with how they serve their audience. The way stories are written today may change, though I doubt the lede will ever disappear. I see it more as a “yes, and” rather than a replacement. We need to arm our reporters and content creators with the right tools so that they can spend more time gathering facts, networking and creating multiple pieces of content. Then, whether it’s video, article, podcast or AI, our audience has what they need.