The end of the op-ed: LA Times uses AI to counteract its own views

Yahoo

16-04-2025

Want to read a contrasting view on the insightful opinion piece you just found? Traditionally, you would have had to find another publication with a different political slant and search for an op-ed on the same topic.
Those days might be numbered as a new feature on the Los Angeles Times website offers an immediate rebuttal to their journalists' work through an AI assistant.
Is this the future of opinion journalism, or have we entered into a parallel universe where even rhetoric is fully supplanted by the machines?
In the LA Times opinion section is an article by Joan C. Williams, a law professor at University of California College of the Law in San Francisco. In the piece, Williams argues that Democrats need to understand the cultural angle of politics if they are ever to understand Donald Trump's 2024 presidential election victory.
'Democrats need to rival Trump's cultural competence at connecting with the working class, which will require changes from progressives,' she writes, as she explains that the left is failing to see the human side of the right's arguments to a broad span of American voters.
Related
Washington Post op-ed writer quits after paper refuses to publish piece critical of Bezos
Volunteer actors rail against Jeff Bezos private performance of 'You Me Bum Bum Train'
It's an insightful and well researched piece that speaks from Williams' academic perspective. Exactly the sort you'd expect from an opinion section.
What's new is a little button labelled 'Insights'. Click that and you're whisked to the bottom of the page where the LA Times gives you an 'AI-generated analysis on Voices content to offer all points of view'.
In the Insights, you are told that Williams' piece 'generally aligns with a centre left point of view' as well as a summary of her points. Most intriguingly, there's a section that gives you opposing views.
There it points out that while Williams thinks Democrats should focusing on uniting cultural theories, strategists like James Carville 'urge Democrats to avoid cultural distractions and concentrate exclusively on economic messaging while others reject the premise that cultural issues are irreconcilable, pointing to Biden-era successes in manufacturing and wage growth'.
On that last point, the AI argues that 'conceding on issues like climate action or LGBTQ+ rights risks demoralizing the Democratic base and ceding ground to far-right narratives'.
This sounds like an interesting point but it's directly contrasted by Williams' own writing where she notes that on inclusive messaging, Democrats 'didn't abandon their vision of full LGBTQ+ rights, but they did shift tactics to engage a wider range of people.'
Whether the AI has fully grasped Williams' argument or not is up for debate, but most significant is the shift away from giving opinion writers a sense of authority on their topic. For many, this is part of a move to placate Trump's administration by LA Times owner Patrick Soon-Shiong.
Soon-Shiong bought the paper in 2018 and last year decided it would not officially endorse Kamala Harris for president. 'If you just have the one side, it's just going to be an echo chamber,' the medical investor said.
Going through the paper's opinion pieces today, the AI provides more opposing views.
LA Times opinion columnist Jackie Calmes has a piece on the parallels between Hannah Arendt's view on Nazi Germany and the current Trump administration.
AI Insights suggests Calmes' piece is 'centre left' and notes that 'Trump's actions, while norm-breaking, lack the systemic violence and ideological coherence of Hitler's regime'.
Similarly, LA Times journalist LZ Granderson has a 'centre left' piece that criticised the Trump administration's removal of LGBTQ+ iconography and education from government bodies as a worrying shift towards un-American ideals of inclusion and openness.
Contrasting this view, the AI suggests: 'Conservative advocates, including groups behind Project 2025, contend that LGBTQ+ visibility policies promote 'toxic normalization' and conflict with traditional family values.'
While the opinion writers of the LA Times are still largely left leaning, the new AI feature is designed to platform both sides of an argument. Exactly as Soon-Shiong wanted.
Back when the owner blocked the Harris endorsement, three of the six people who researched and wrote their editorials, including editorials editor Mariel Garza, resigned in protest. The other three have since left, with the last holdout, Carla Hall, exiting after writing a last column that ran on 30 March about homeless people she met while covering the issue.
Soon-Shiong's decision was mirrored by Washington Post owner Jeff Bezos, who decided the newspaper would not back a presidential candidate.
Last month, Ruth Marcus, who worked for the Washington Post since 1984, stood down after a row over the paper refusing to run an opinion piece that was critical of Bezos.
She was the latest to leave the Post after Bezos announced that the paper's opinion columns would no longer be a free-for-all for individual author's thoughts. Instead, it would focus on two specific themes: 'personal liberties and free markets'.
Between Soon-Shiong and Bezos, there is a clear move from the billionaire-owned US media landscape to cosy up to the Trump administration. It continues a pattern seen by the major tech and social media owners, including the Bezos-founded Amazon.
Concerningly, it also highlights AI-obsessed owners' cynical attitude towards the work of journalists.
'I think it could be offensive both to readers ... and the writers themselves who object to being categorized in simple and not necessarily helpful terms,' Garza, the ex-editorials editor said.
'The idea of having a bias meter just in and of itself is kind of an insult to intelligence and I've always thought that the readers of the opinion page were really smart."