Patrick Soon-Shiong, the erratic billionaire owner of the Los Angeles Times, is touting the single worst idea ever seriously contemplated by a newspaper executive and trust me that is saying something.
In October, Soon-Shiong blocked his editorial board from endorsing Kamala Harris, sparking a wave of resignations. That was the first indication of his sharp turn from reasonable, apparently liberal owner to MAGA revolutionary.
Now, he is having the tech team at his biotech company create a “bias meter” intended to expose the bias of the reporters and columnists who work for him.
And he is also having them create a button that generates two AI-written versions of “both sides” of the story — one of which would presumably be a Trump-friendly version of the news.
He announced these moves in a podcast interview with Scott Jennings, the right-wing provocateur who he just hired for his editorial board. His explanation was that he recently concluded that the Times is “an echo chamber, not a trusted source.”
His proposed solutions, however, suggest he has no idea how journalism actually works.
The bias meter is utterly ridiculous. The results would be specious and all it would accomplish at best would be to undermine his own staff.
But the button is the real killer. Leave aside that there aren’t two and only two sides to a lot of topics, how do you create “both sides” versions of a story based on reported facts? What’s the “other side” of a fact? Do you have the AI create a version in Trumpese? Full of lies?
It sounds like one of the two “button” versions would be true, the other would be false, and they’d both be written by AI instead of human. It’s perverse.
The incredible irony here is that Soon-Shiong insists that this is a way of rebuilding trust. It would do the opposite, while destroying the institution he bought in 2018, ostensibly to save it.
To be clear: AI cannot be trusted to do serious journalism. It can create the reasonable facsimile of a news story from structured data, like stock market values; it can summarize an article fairly effectively. But no reputable news organization would trust AI to write complicated stories on its own. Previous attempts to do so have ended in disaster.
Then there’s the small matter that, especially in this era of disinformation, both-sides reporting is the worst way to deliver the news. We need reporters to distinguish between the truth and lies, not simply present both and let the reader decide.
Soon-Shiong’s plan fundamentally misunderstands what real, human journalists do. It makes a mockery of the process reporters go through, which is trying to determine the truth and subjecting it to verification.
And, if he pursues his plan, it is sure to provoke a widespread revolt among his staff.
The LA Times Guild has already issued an angry statement, criticizing Soon-Shiong for suggesting his staff “harbors bias, without offering evidence or examples.”
And, the statement said, the guild’s members “will firmly guard against any effort to improperly or unfairly alter our reporting.”
I don’t know what happened to Soon-Shiong – whether it was a business decision, a falling-out, or a stroke. But he’s really gone around the bend.
Oliver Darcy, who writes the Status newsletter, had an extraordinary interview with Soon-Shiong last month, in which the billionaire asserted that it is an “opinion,” not a matter of fact, that Donald Trump lies at a higher rate than most other politicians. He accused Darcy of “bias” and said “This is really what I think is the matter with the country.”
Combine all this with Jeff Bezos’s recent moves at the Washington Post and the message is clear: Billionaires cannot be trusted to preserve the legacy of our great journalistic institutions. Indeed, they might well destroy them.
The next he’ll require reporters to take lie detector tests as they write a story. Then he’ll know in real time when a story is going off the rails, and he can edit them when he’s not re-writing headlines.
Patrick Soon-Shiong is an exemplar of what is wrong with the news business. It is a business but reporting facts does not seem to factor into the business model. Various corporate interests and oligarchs have positioned themselves at the faucet of information and they mean to control what their customers consume. I am beginning to think that I have a naive belief that reporters should report what actually happens. It seems the owners of our news media would rather abstract away from the facts of the day. Often the last thing they want is for people to see what actually happens. They would rather cut information up. Put it in snippets. Report in incomplete parts and then have their reporters fill in the blanks with narratives. The less complete the picture, the easier it is to let the talking heads take center stage and drive the consumer to whatever conclusions are desired. If American’s were trained in critical reasoning this would be harder to do, but alas, Americans do not tend to get trained in critical reasoning.
Anyone who trusts AI deserves what they get. Any company that uses generative AI has forsaken the right to have anything they say taken seriously.
It’s terrible that my relationship to the la times (60) years may end because of this horrible decision by the clueless owner.
It is indeed terribly sad, really.