G/O Media Gives Another Crash Course On Perils Of Replacing Human Journalists With Half-Baked ‘AI’
While recent evolutions in AI" have netted some profoundly interesting advancements in creativity and productivity, its early implementation in journalism has been a sloppy mess thanks to some decidedly human-based problems: namely greed, incompetence, and laziness.
If you remember, the cheapskates over at Red Ventures implemented AI over at CNET without telling anybody. The result: articles rife withaccuracyproblemsandplagiarism. Of the 77 articles published, more thanhalfhad significant errors. It ultimately cost them more to have humans editors come in and fix the mistakes than the money they'd actually saved.
Similar efforts are afoot over at G/O Media (owner of Gizmodo), which has eagerly been replacing staff with half-baked AI," despite very clear and repeated promises they wouldn't be doing that. Last week,G/O Media shut down the blog's Spanish-language site,Gizmodo en Espanol. Human writers were replaced with an AI-powered translation tool that would basically just translate existing Gizmodo content.
It... did not go well. There were numerous instances where the AI couldn't even translate entire articles competently (the whole point), switching between Spanish and English mid article. There were numerous other issues where the AI screwed up basic formatting and punctuation:
Take thetranslated versionof arecent Star Wars recap, which in English was titled Ahsoka'Had Another Major Cameo You Definitely Missed." The AI seemingly doesn't quite understand how to populate the page, cutting off paragraphs at strange points and blundering punctuation. One paragraph even includes a bit of HTML thatdefinitelywasn't supposed to make it onto the page, reading: <caption>Kanan!</caption>"
One one hand, translation is obviously a useful application of LLMs. On the other hand, the kind of media executives rushing to replace humans with this stuff (without ensuring it works first) are genuinely not good at their jobs. They don't understand tech or the media environment they operate in, have routinely bled top-shelf human talent, and primarily see AI as a way to cut corners and screw already underpaid labor.
They don't care that a robot can't really determine the subtle, localized nuances Spanish-speaking Gizmodo readers might be interested in. They don't care that the end product is of a decidedly lower quality (or they'd pause implementation until they knew what they were doing). They see AI" as a cost-cutting shortcut and a useful bludgeon against labor, especially the unionizing variety.
AI will absolutely be useful in journalism in a myriad of ways. And it will certainly replace many of the lower-hanging, underpaid staff writing more remedial content. But the technology is being layered on top of a very broken U.S. journalism and media industry run by terrible and incompetent human beings. Folks with an obvious disdain for their human staff who prioritize clickbait bullshit over quality journalism.
While AI technology will improve, there's very little indication the media executives implementing it will. For many of these fail-upward media executives, the goal is to use AI to effectively create a giant automated ouroboros of clickbait engagement porn that basically shits money. Product quality, happy employees, or the sanctity of journalism as a public service simply never enters into it.