The "More Human" Journalist

In March 2014, Ezra Klein launched Vox with a promise to “explain the news.”[1] The pitch was simple: in a world drowning in information, what readers needed was context. For nearly a decade, explainer journalism thrived. Vox’s model — combining accessible writing with deep research — became the template for a generation of digital publishers.

That era is ending.

Today, when someone needs to understand a complex policy debate or contextualise breaking news, they increasingly turn to ChatGPT. The AI synthesises thousands of sources, tailors the explanation to exactly what you need, and delivers it in seconds. A 2024 Reuters Institute survey found that only 47% of journalism leaders express confidence about the profession’s future — the lowest level in a decade.[2] The challenge is existential: if your journalism can be replicated by an algorithm, it has no economic future.

This moment demands a reckoning. Journalism must return to what algorithms cannot touch: the irreducibly human. The visceral experience of being there. The texture of a voice caught in uncertainty. The smell of a place that data cannot convey. The 30,000 sensory inputs we process every second, of which only 10 to 15 reach conscious awareness — but all of which shape how we understand the world.

Johnny Harris left Vox in 2020 to become an independent journalist on YouTube. Today, he has more than 7.5 million subscribers.[3] Cleo Abram made a similar leap in 2022, growing to more than 7.5 million as well.[4] These aren’t outliers — they’re the leading edge of a structural shift.

“There’s a lot of people that think that journalism is a status,” Harris has said. “But I think journalism is a practice.”[5] That distinction matters. When journalism was a status — conferred by institutions like The New York Times or The Washington Post — individual journalists derived authority from their affiliation. Now, with trust in U.S. media at 28%,[6] authority accrues to individuals who build direct relationships with audiences.

The economics follow. A journalist with several million YouTube subscribers can generate revenue that rivals legacy institutions. They control their narrative voice, choose their stories, maintain direct audience connection. Harris’s venture, NewPress, exemplifies the emerging model: a collective of journalist-creators sharing infrastructure while maintaining individual brands.[7]

For traditional newsrooms, this raises an uncomfortable question: why would a journalist with the option to build their own platform accept institutional constraints?

Increasingly, they won’t.

But if institutions crumble, what fills the gap?

Since 2005, the United States has lost more than 2,900 newspapers, with over 200 counties now classified as “news deserts.”[8] But information doesn’t disappear — it reorganises.

In communities where newspapers have shuttered, networks emerge organically. WhatsApp groups that started for organising school pickups become disaster response systems. Church leaders announce community events. Neighbourhood Facebook groups share news about road closures and local government decisions.[9] These aren’t journalism in the traditional sense, but they fulfil many of journalism’s functions: distributing information, coordinating community response, providing early warning.

Digital-native Latinx news outlets show what adaptation looks like in practice. Rather than pulling audiences to websites, they meet communities where they already gather — on messaging apps.[10] The power dynamic inverts. For decades, journalists controlled narrative by controlling distribution. That monopoly is gone.

When information flows through community networks, a story isn’t told about a community — it’s told by the community. The journalist’s role shifts: less gatekeeper, more facilitator. Research shows that communities with strong information networks demonstrate greater civic engagement and more effective crisis response.[11] Journalists who help build those networks, even when it means sharing authority, create something no algorithm produces: trust built over time.

Communities can distribute information. They can coordinate action. But can they create the kind of understanding that changes how someone sees the world?

That requires something else entirely.

In June 2011, I started a journey across Africa with a question I couldn’t stop thinking about: what if the stories that connect us matter more than the stories that separate us?

I had been working in different countries across the continent for some years by then, and a few months earlier I’d done a small experiment in Rwanda — walking up to people on the street and asking them, simply, “What makes you happy?” When I played that footage during presentations later, something unexpected happened. People didn’t respond to what I considered the big stories — the ones about war, about poverty, the important ones. They responded to the universal story. To someone in Butare saying, “I’m happy when I play music,” or “I’m happy when I’m together with my friends.” In that moment, the distance collapsed. Suddenly there was connection.

Out of that, I developed ten questions about life — What makes you happy? What is most important to you? What are your biggest struggles? What do you dream about? — and travelled from Cape Town to Cairo over six months, using only public transport, recording hundreds of interviews along the way. The idea was simple: start with what unites, and the rest follows.

Those connecting stories became more than content. They became a way in — a way to open our own minds, to think differently. To listen before reacting. I still ask people those questions when I interview them, years later. They have a power in them that reveals something. Everyone has dreams. Everyone struggles. Everyone can name what makes them happy.

I think about this now when people ask whether AI will replace journalism. The question misses something fundamental about how stories work.

A story is a bridge between a lived experience and the reader. AI can build the visible structure of that bridge — the sentences, the arc, the argument. It’s getting remarkably good at it. But the bridge needs more than structure. It needs the 90% that remains invisible: the subtle pause before someone answers a difficult question, the way they start thinking, the smoke from cooking fires burning in your eyes at a dumpsite in Nairobi, the specific weight of silence in a room where someone has just told you something they’ve never told anyone.

AI has access to the 10% that gets captured in some form — the compressed output. It doesn’t have the 90% that led to it. And here’s what matters: stories live from two things at once. They live from the richness of sensory detail and experience lying just under the surface — shaping how we feel about what we write, what angle we take, what we notice. And they live from the space they leave for the reader — the room to interpret, to bring their own memories, to connect.

A story that tried to capture everything would be overwhelming and purposeless. The bridge works precisely because it reduces: it takes the immense complexity of lived experience and distils it into something crossable. But compression requires knowing what to keep — which details carry the emotional weight of the whole, which pause holds a life story in it. That judgment only comes from having stood on both sides of the river. Without the 90%, you don’t know which 10% matters.

I learned this viscerally, years earlier, making a documentary about climate change in Kenya. I travelled to the north, to drought areas where it hadn’t rained for a year and a half. I remember the animal skeletons lying in the desert — cattle, mostly, that had simply run out of water. People who were already impossibly thin. The reality of climate change was stark and immediate and physical. And then I came home to Denmark carrying all of it — the heat, the silence, the thin bodies against dry earth. But I couldn’t transfer it. People around me had the same data I had. They knew the projections, the temperature curves, the drought statistics. What they didn’t have was what it felt like to stand there. I had crossed the river. They hadn’t. And no amount of data could build that bridge for them. And I had focused to much in the data.

That gap — between data and experience — is where journalism either lives or dies.

When I filmed Wanjira Wanjiru in Mathare, one of Nairobi’s largest informal settlements, she told me about her nineteen-year-old brother — a promising footballer, shot in the back by police during the 2007 post-election violence. She described going to the mortuary, reading the report that confirmed it was a police shooting. That pause before she spoke — the years of grief compressed into a single breath — cannot be synthesised from data.

When war correspondent Mstyslav Chernov filmed inside the Mariupol maternity hospital as Russian bombs fell, the shaky camera wasn’t a technical flaw. It was the signature of a body under threat. His film won the 2024 Academy Award for Best Documentary Feature Film.[12]

There’s a neurological reason this works. Mirror neurons fire both when you perform an action and when you watch someone else perform it.[13] When a journalist describes the slight catch in someone’s voice, the way their hands trembled, the long pause before answering — these sensory details activate the reader’s neural architecture for empathy. Neuroscientists have shown that during storytelling, brain patterns can synchronise between storyteller and listener — a phenomenon researchers call neural coupling. Empathy at the neurological level.

This is why immersion journalism creates such powerful narrative. Ted Conover didn’t just write about prison guards; he became one for a year at Sing Sing.[14] David Finkel didn’t interview soldiers returning from Iraq; he moved into their homes and documented their struggles over years.[15] These stories succeed because they offer embodied knowledge: what it feels like to do night shift on a cellblock, the exhaustion of hypervigilance, the specific texture of fear.

I felt this distinction sharply just weeks ago in Uganda, filming a documentary about flooding in a community called Kitintale. I was working with Drake, a community journalist — sharp, skilled, very much a facts-first reporter. He got the information efficiently: the flood levels, the damage reports, the number of homes destroyed. All accurate. All important. But sitting with a man whose house had been completely washed away — who had pulled his son from rising water in the dark — I realised the story needed something Drake wasn’t reaching for. The pause. The weight of what it means to fill your home with soil after every flood, knowing the next one will wash it away again. The facts told you what happened. Being there told you what it meant.

AI can summarise research about empathy. It cannot make you feel the crampedness of a shack in Korogocho or hear the uncertainty in a refugee’s voice when they describe why they left home. That sensory specificity — the details that build the bridge between one person’s experience and another’s understanding — remains the province of humans who were actually there.

If embodied experience is journalism’s survival mechanism, then the industry is measuring the wrong things.

Newsrooms, desperate for traffic as advertising revenue collapsed, began making editorial decisions based on real-time analytics. Research by sociologist Angèle Christin found that journalists internalised these metrics, writing headlines and framing stories to optimise for clicks — even when they knew it degraded quality.[16] The attention economy created a race to the bottom: whoever triggered the strongest emotional reaction won the algorithmic lottery.

But attention is extractive. It depletes rather than nourishes. Media theorist Yves Citton argues for an alternative: the curiosity economy.[17] Where attention is scarce and must be captured, curiosity is expansive — every answer generates new questions. An attention-optimised headline reads “You Won’t Believe What This Politician Just Said.” A curiosity-optimised approach asks, “Why Are Infrastructure Bills So Hard to Pass?” The former extracts a click. The latter invites thinking.

The trap is self-reinforcing. As journalism becomes more algorithm-optimised, it becomes more replaceable by algorithms. Why pay for journalism that delivers the same emotional triggers as free AI-generated content?

Breaking the cycle means measuring different things. Instead of page views, track how many readers finish the story. Instead of social shares, measure whether readers take action. Instead of virality, optimise for what Citton calls “curiosity generation” — whether a story makes readers ask new questions rather than simply consuming an answer.[17]

Danish news startup Zetland starts each morning by singing together as a staff — a ritual designed to centre human connection over metrics.[18] They measure success not by traffic but by whether readers feel their lives are enriched. In a market flooded with algorithmic content, genuine human connection becomes the differentiator.

The stories that break through don’t lead with statistics or policy frameworks. They start with a scene: the smell of a place, a specific moment, a physical detail that anchors you in lived experience. Journalists trained to write inverted pyramids — most important information first — must unlearn that reflex. Stories that create immersion build slowly, layering detail and withholding resolution. They’re structured more like fiction: establishing character, building tension, pulling readers forward through questions they need answered.

This requires time. Deep reporting. Multiple interviews. The patience to catch contradictions that reveal complexity. Observation that captures not just what people say but how they say it. AI excels at transcription, translation, research synthesis — infrastructure that frees journalists to do work only humans can do.[19] The core competency remains human: the judgment to know what matters and the empathy to understand why.

In his meditation on walking across continents, journalist Paul Salopek describes the rhythm of moving at five kilometres per hour — the speed of human feet, the pace at which genuine observation becomes possible.[20] “I think that this sense of well-being that comes with timelessness, the sense of being at peace — it must be very, very old,” he reflects.

I think about pace often. Some of my clearest memories from years of fieldwork come from the mornings — those first hours before the day fully arrives. Walking through Khayelitsha, a large township in South Africa, at dawn during the Humans Project, the settlement quiet and misty, the sun not yet up. The night had been tense — people banging on doors, a sense of the neighbourhood’s edges. But the morning had a different texture entirely. Soft light. Schoolchildren appearing on paths. The world waking up slowly. I was walking with a young guy to his school, through narrow lanes between tin shack houses, past lines of white clothes hanging in still air. Everything felt airy and open, nothing like the daytime density. I remember thinking: this is when you actually see a place. When it’s more than you expect.

Those mornings taught me something about journalism that no methodology course ever did. The real story rarely arrives on schedule. It surfaces in the gaps — in the walk to school, in the kitchen before the interview starts, in the silence after someone finishes speaking and you have the discipline not to fill it. The 90% that informs the bridge doesn’t announce itself. You have to be present long enough for it to appear.

There is something in Salopek’s slowness worth taking seriously. AI is a tool, and tools properly used expand what is possible. But journalism was something before it became an industry optimised for speed: the practice of paying attention. Of being present long enough to notice what the official story leaves out. Of staying until the second version of events surfaces, the one that contradicts the first.

The journalist who survives the age of algorithms won’t be the one with the best AI tools. It will be the one who still knows what it feels like to sit with someone in silence before they are ready to speak — and who understands why that silence is worth more than any summary.

Share

[1] Vox Media. (2014). “Nine questions about Vox.” Vox. Retrieved from https://www.vox.com/2014/3/28/5559144/nine-questions-about-vox

[2] Newman, N. (2024). “Journalism, Media, and Technology Trends and Predictions 2024.” Reuters Institute for the Study of Journalism. Retrieved from https://inpublishing.co.uk/articles/ai-and-journalism-publishers-brace-for-a-new-wave-of-disruption-22986

[3] SPEAKRJ. (2025). “Johnny Harris YouTube Net Worth and Estimated Income Stats.” Retrieved from https://www.speakrj.com/audit/report/UCmGSJVG3mCRXVOP4yZrU1Dw/youtube/income-stats

[4] Created. (2024). “Why Every Cleo Abram Video Goes Viral (Interview).” Retrieved from

[5] Colin and Samir. (2025). “An Honest Conversation with Johnny Harris.” Retrieved from

[6] Gallup. (2024). “Trust in Media at New Low of 28% in U.S.” Retrieved from https://news.gallup.com/poll/695762/trust-media-new-low.aspx

[7] Palmeri, T. (2025). “YouTube star Johnny Harris on his Growing Media Empire.” The Red Letter. Retrieved from

[8] Franklin, T. (2025). “The State of Local News 2025.” Medill Local News Initiative. Retrieved from https://localnewsinitiative.northwestern.edu/projects/state-of-local-news/2025/

[9] Robertson, K. (2024). “Spurned by Social Media, Publishers Chase Readers on WhatsApp.” The New York Times. Retrieved from https://www.nytimes.com/2024/09/24/business/media/whatsapp-news-outlets.html

[10] Chacón, L. C. (2024). “¿Qué pasa with American News Media? How Digital-Native Latinx News Serves Community Information Needs Using Messaging Apps.” Media Policy for an Informed Citizenry. Retrieved from https://mediawell.ssrc.org/articles/media-policy-for-an-informed-citizenry-revisiting-the-information-needs-of-communities-for-democracy-in-crisis/?pdf=78634

[11] Knight Foundation. (2023). “Knight Communities Healthy News & Information Ecosystems 2023.” Retrieved from https://knightfoundation.org/reports/knight-communities-healthy-news-information-ecosystems-2023/

[12] PBS. (2025). “20 Days in Mariupol.” Frontline. Retrieved from https://www.pbs.org/wgbh/frontline/documentary/20-days-in-mariupol/

[13] Rizzolatti, G., & Fabbri-Destro, M. (2022). “Mirror neurons 30 years later: implications and applications.” Trends in Cognitive Sciences. Retrieved from https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(22)00134-6

[14] Conover, T. (2016). Immersion: A Writer’s Guide to Going Deep. University of Chicago Press. Retrieved from https://tedconover.com/book/immersion/

[15] Finkel, D. (2014). “David Finkel and the Art of Immersion Reporting.” Nieman Reports. Retrieved from https://niemanreports.org/david-finkel-and-the-art-of-immersion-reporting/

[16] Christin, A. (2020). Metrics at Work: Journalism and the Contested Meaning of Algorithms. Princeton University Press. Retrieved from https://press.princeton.edu/books/ebook/9780691200002/metrics-at-work

[17] Citton, Y. (2022). “From the Attention Economy to the Curiosity Economy.” Electra Magazine. Retrieved from https://electramagazine.fundacaoedp.pt/en/editions/issue-23/attention-economy-curiosity-economy

[18] WAN-IFRA. (2024). “How thinking about jazz, milk and high quality red wine have informed Zetland’s success.” Retrieved from https://wan-ifra.org/2024/07/how-thinking-about-jazz-milk-and-high-quality-red-wine-have-informed-zetlands-success/

[19] EBU. (2025). “EBU News Report 2025: Leading Newsrooms in the Age of Generative AI.” Retrieved from https://www.ebu.ch/files/live/sites/ebu/files/Publications/Reports/open/EBU_News_report_2025_Leading Newsrooms_AI.pdf

[20] Salopek, P. (2024). “A Path Older Than Memory.” Emergence Magazine. Retrieved from https://emergencemagazine.org/interview/a-path-older-than-memory/