At 21, David Pakman started a little Massachusetts community radio talk program. While the young broadcaster got his show syndicated on a few public radio stations, it was a YouTube channel he began in 2009, “The David Pakman Show,” that opened up his progressive political commentary to a whole new digital audience. The show has since amassed 353,000 subscribers, and roughly half its revenue now comes from the ads that play before his videos. He earns enough to produce the show full time and pay a lean staff.

Or, at least, he used to. Last month YouTube announced abrupt, vague changes to its automated processes for placing ads across the platform. Ads on Pakman’s YouTube channel evaporated, dropping to as little as 6 cents a day, and forcing him to set up a crowdfunding page to help cover $20,000 a month in operating costs.

“This is an existential threat to the show,” Pakman said. “We need that money.”

Since its 2005 debut with the slogan “Broadcast Yourself,” YouTube has positioned itself as a place where any people with camera phones can make a career of their creativity and thrive free of the grip of corporate media gatekeepers. But to share in the advertising wealth a user base of more than 1 billion can provide, independent producers like Pakman must satisfy the demands of YouTube’s unfeeling, opaque and shifting algorithms.

The architecture of the internet has tremendous influence over what is made, and what is seen; algorithms influence what content spreads further on Facebook and turns up on top of Google searches. YouTube’s process for mechanically pulling ads from videos is particularly concerning, because it takes aim at whole topics of conversation that could be perceived as potentially offensive to advertisers, and because it so often misfires. It risks suppressing political commentary and jokes. It puts the wild, independent internet in danger of becoming more boring than TV.

YouTube’s most serious ad change yet came in the wake of reports from The Times of London and The Wall Street Journal that ads were appearing on YouTube videos that espoused extremism and hate speech. When major advertisers like AT&T and Johnson & Johnson withdrew their spots, YouTube announced that it would try to make the site more palatable to advertisers by “taking a tougher stance on hateful, offensive and derogatory content.”

But according to YouTube creators, that shift has also punished video makers who bear no resemblance to terrorist sympathizers and racists. YouTube’s comedians, political commentators and experts on subjects from military arms to video games have reported being squeezed by the ad shake-up — often in videos they’ve posted to YouTube. On the site, it’s known as “the adpocalypse.”

The topics of Pakman’s videos are no more controversial than the programming typically found on CNN or the local news. In fact, because his show is also broadcast over the radio, it adheres strictly to FCC content rules. But the show takes pride in its independence from corporate ownership. “I have no boss above me,” Pakman said, and because of YouTube’s automated ad systems, “I’ve never had any contact with advertisers, so it’s impossible for me to ‘sell out’ to satisfy them.”

Instead, he’s subject to the whims of the algorithm. To rein in its sprawling video empire — 400 hours of video are uploaded to the platform every minute — YouTube uses machine learning systems that can’t always discern context, or distinguish commentary or humor from hate speech. That limitation means that YouTube routinely pulls ads from content deemed “not advertiser-friendly.” That includes depictions of violence or drug use, “sexual humor” and “controversial or sensitive subjects,” including war and natural disasters. YouTube has previously blocked ads on Pakman’s news videos referring to ISIS church bombings and the assassination of a Russian diplomat. YouTube creators can appeal the decision to a human who will watch the video, but the recent changes have upended the usual processes, often leaving creators without an official channel for appeal.

Recent ad changes have hit some of YouTube’s most popular channels: The comedy channel h3h3Productions has seen YouTube’s system grow stricter in recent weeks, while the Military Arms Channel, which tests and reviews firearms, has registered a complete drop in ad revenue before rebounding to 25 percent of its usual income. Felix Kjellberg, a wildly successful comedy and video game vlogger known as PewDiePie, lost development deals with YouTube and the Disney-owned Maker Studios after The Wall Street Journal reported on a proliferation of Hitler references in his videos. (Kjellberg apologized for what he called jokes that went too far.) But he’s also seen programmatic ads yanked from a portion of his videos, and he’s become a symbol for the resistance to the “adpocalypse.” He’s taken to churning out satirical videos presenting himself as the brand-safe character “Family-Friendly Felix.”

Last month’s changes signaled that YouTube’s ad rules had become even stricter and less clear. YouTube announced that it was leading advertisers toward “content that meets a higher level of brand safety.” YouTube gave creators no further hints on what that means, just issuing a blanket message telling them they may experience “fluctuations” in revenue as the new systems are “fine-tuned.”

“It’s getting so bad that you can’t even speak your mind or be honest without fear of losing money and being not ‘brand-friendly,’” said Ethan Klein, creator of h3h3. “YouTube is on the fast track to becoming Disney vloggers: beautiful young people that wouldn’t say anything controversial and are always happy.”

This is not all the algorithm’s fault. People create these systems, and they are sensitive to bad press and skittish advertisers. The YouTube ad crisis — and the company’s response — also speaks to a persistent public misunderstanding of the worth of digital creators. The mainstream media barely engages with YouTube videos as an artistic product in the way it does traditional television and film. Coverage is focused on how much money YouTube stars make, how improbably famous they are among teenagers and, now, on the small number of racist and extremist videos that have slipped through the cracks of the ad system.

Jamie Byrne, director of creators and enterprise at YouTube, said concerned companies had requested tougher controls to keep their ad dollars flowing. “For creators to flourish on our platform, we need an incredibly strong advertising community engaged on YouTube as well,” he said. He hopes that as the ad systems learn to decipher context, and advertisers relax, creators will see greater returns.

Media power has consolidated in the past several decades, with a smaller crew of billionaires controlling journalism and entertainment. The internet was supposed to offer an opportunity for a diverse group of upstarts to challenge the corporate structure. But the same consolidation of power has happened online. Independent blogs have been shut down or snapped up by bigger companies. YouTube has long been owned by Google, which has gobbled up a greater share of online ad revenue in recent years. Google and Facebook are so dominant that they form a practical duopoly over digital advertising. Meanwhile, YouTube is making a run at Netflix with its original-content subscription service, YouTube Red, and going after television with YouTube TV, which allows subscribers to stream TV channels online for $35 a month.

All of that means that new media creators hoping to make a living online need to play by YouTube’s rules, and steer clear of anything “potentially objectionable” — not to real people, who might actually be offended, but to robots. If YouTube wants to fulfill its promise of an online environment where independent creators can make interesting work, it will find a way to scrub ads from truly vile content without penalizing the merely controversial.

© 2017 The New York Times Company