A former alt-right YouTuber details how he tailored videos for the “echo chamber”, including focusing squarely on confrontations to drive views on YouTube
Focus on conflict. Feed the algorithm. Make sure whatever you produce reinforces a narrative. Don't worry if it is true. Tweets: @nytimesbusiness , @markhurst , @jgreenblattadl , @frankpasquale , @nytmedia , @anthony , and @clearing_fog Tweets: @nytimesbusiness : Over the more than two years Caolan Robertson helped produce and publish videos for right-wing YouTube personalities, he learned how making clever edits and focusing on confrontation could help draw millions of views. https://www.nytimes.com/... Mark Hurst / @markhurst : “ ‘We realized that if we wanted a future on YouTube, it had to be driven by confrontation.’... the longer someone watches, the more extreme the videos can become.” Google is not trustworthy. https://www.nytimes.com/... Jonathan Greenblatt / @jgreenblattadl : Powerful piece by @CadeMetz exploring how #extremists knowingly exploited @YouTube's algorithm, spreading #hate and #harassment widely. https://www.nytimes.com/... Frank Pasquale / @frankpasquale : “It can create these very radical people who are like gurus,” said Guillaume Chaslot, a former YouTube engineer who has been critical of the way the company's algorithms pushed people to extreme content. “In terms of watch time, a guru is wonderful.” https://www.nytimes.com/... @nytmedia : Caolan Robertson learned how making clever edits and focusing on confrontation could help draw millions of views on YouTube and other services. He also learned how YouTube's recommendation algorithm often nudged people toward extreme videos. https://www.nytimes.com/... Anthony DeRosa / @anthony : “It appears in the videos that we are just trying to figure out what is going on, gather information, understand people,” Mr. Robertson said. “But really, we were trying to find the most incendiary way of making them mad.” https://www.nytimes.com/... ClearingTheFog / @clearing_fog : Like Facebook, YouTube's algorithms introduce and recruit people to right wing extremist ideologies. Google and Facebook have been exponentially growing extremist groups for years, and refusing to admit it when questioned. https://twitter.com/...
Context & Ripple Effects
Caolan Robertson's account is an inside-source sequel to a years-long arc of coverage: earlier reporting traced how teens were radicalized by YouTube's recommendation algorithm and called the platform one of the most powerful radicalizing tools of this century. What the algorithm-blame debate lacked was testimony from the production side — and Robertson, who spent over two years producing videos for right-wing personalities, supplies it: clever edits, engineered confrontations, and narratives built for reinforcement rather than accuracy.
His account lands in the middle of an unresolved dispute. A 2019 report argued right-wing proliferation on YouTube was driven by supply and demand rather than the recommendation algorithm — and Robertson's confession is effectively a supply-side exhibit, showing creators deliberately feeding the demand the platform monetizes. It also revives the enforcement question YouTube has never cleanly answered: whether it will punish its own stars.
First-order effects
- Robertson's testimony reframes the radicalization debate around creator intent: the confrontational edits and echo-chamber framing were deliberate production choices by the right-wing personalities he worked for, not just algorithmic accidents — direct ammunition for critics of those channels and their advertisers' calculus.
Second-order effects
- YouTube faces renewed pressure on its unresolved enforcement dilemma — curbing misinformation likely means punishing some of its own stars, a step it has not shown it can broadly take — while the supply-and-demand findings give the platform a counterargument that blame is shared with creators and audiences.
Third-order effects
- If the pattern holds, platform accountability will shift from policing individual videos to scrutinizing the production playbook itself — the repeatable formula of conflict-first editing and narrative reinforcement — forcing regulators and advertisers to decide whether engagement-optimized content design is a moderation problem or a business-model one.
The trend: The radicalization story is moving from 'the algorithm did it' to a co-production model in which creators deliberately engineer conflict for engagement and platforms profit from the demand that engineering creates.