TikTok's recommendation engine has always been a niche-amplification machine — it finds pockets of genuine enthusiasm and blasts them into the mainstream feed. TrainTok is the latest proof of concept. Posts tagged #UKTrains have jumped 247% in three months, powered by a mix of traditional trainspotters, transport workers sharing insider knowledge, and educational creators explaining why DLR carriages wobble. The most interesting strand here isn't the trainspotters — Francis Bourgeois proved that worked back in 2021. It's the transport workers. Shamal Cain, a TfL employee posting as Tube Bellingham, noticed that millions of daily commuters didn't know about the TfL journey planner or how penalty fares worked. That's not entertainment content — it's a public information service delivered through a format people actually consume. When followers started DMing him about alternative routes during service disruptions, he'd effectively become an informal customer support channel running on personal time. Nathan Pinnock-James (Network Nathan), a 24-year-old rail enthusiast, represents the other engine: pure curiosity content. His videos on trainspotting locations, station openings, and the physics of train wheels hit the sweet spot TikTok's algorithm rewards — specific enough to feel like insider knowledge, visual enough to hold a scroll, and short enough to finish. "TikTok has really opened the door to really niche, specific topics," he says, and the data backs him up: UK history searches are up 50%, National Trust posts have doubled. The article buries the most important paragraph near the bottom. AI-generated train content — including fabricated clips of high-speed trains and passengers falling from tube carriages — is also contributing to the growth numbers. TikTok disclosed in late 2025 that more than 1 billion AI-generated videos exist on the platform. That 247% surge in #UKTrains isn't purely organic human enthusiasm; some unknown fraction is synthetic slop riding the same hashtag. TikTok's framing of the trend, delivered during an annual trend report at its UK headquarters, is predictably self-serving. Yazmin How, head of emerging communities, calls it "authentic, educational storytelling" — but the platform has no obvious incentive to distinguish a TfL worker's genuinely useful commuter advice from AI-generated engagement bait, because both produce the watch-time metrics that matter. What's actually happening here is a three-layer phenomenon dressed up as one heartwarming story. Layer one: real human experts sharing real knowledge (Cain, Pinnock-James). Layer two: algorithmic amplification turning a niche hobby into a trend category the platform can package in reports. Layer three: AI-generated content inflating the numbers. The Guardian covers all three but treats them as a single cheerful narrative about British eccentricity going digital. The reader deserves to see the seams.
