June 1, 2017

"New York Times picks an AI moderator over a Public Editor."

That's the headline at Engadget in a piece by Andrew Tarantola about the elimination of the "Public Editor" position at the NYT. I've already said a few words — at the end of this earlier post — about this change at the NYT, but I want to continue the discussion here.

What's the "AI moderator"? The Engadget headline is very confusing since it made me think the Times had artificial intelligence that could substitute for the role the Public Editor had played, which was to monitor the journalism in the newspaper. But, reading on, I see that the AI is about culling the comments:
The NYT's commenting system is powered by Google Jigsaw's Conversation AI, a neural network that has been trained to find and flag trolling, hate speech and gratuitous shitposts in the paper's online comments sections. However, the system is currently only working on around ten percent of the sites articles. With Wednesday's announcement, the program will be expanded to nearly all of the publication's articles. "This expansion," Sulzberger Jr. wrote [in a staff memo], "marks a sea change in our ability to serve our readers."
So in addition to getting the Public Editor out of the way, the NYT is filtering out the contributions of readers who might push back against the distortions, omissions, and fake news. That's pretty interesting, considering that Sulzberger's staff memo also justified eliminating the Public Editor on the ground that readers were performing the role of keeping the NYT principled and honest.

From the memo (which appears in full at the link):
The responsibility of the public editor – to serve as the reader's representative – has outgrown that one office. Our business requires that we must all seek to hold ourselves accountable to our readers. When our audience has questions or concerns, whether about current events or our coverage decisions, we must answer them ourselves....
Wow! Did you see the sleight of hand? The public editor was "the reader's representative," but now "we... all" should answer to the readers directly. Then who is representing the readers? You have no intermediary. You have the unrepresented readers, trying to make their questions and concerns heard with no surrogate on the inside, and you're also systematically cutting them off, using automation to distance them even further.
The public editor position, created in the aftermath of a grave journalistic scandal, played a crucial part in rebuilding our readers' trusts by acting as our in-house watchdog....
So you got that taken care of? You think we trust you now? Was it all only about mopping up after the scandal, and that's far enough behind you now? What matters isn't just whether we trust you — and obviously not all of us do — but whether you are actually doing an excellent job according to the best principles of journalism. But I hear you saying, that's good enough trust for now. Or even: The public editor brought some trust but also mistrust, and when trust/mistrust balance cut the wrong way, we ended it.

Here's the pitch from Sulzberger that times have changed and the internet is all the pushback needed to keep the NYT honest and principled:
[T]oday, our followers on social media and our readers across the internet have come together to collectively serve as a modern watchdog, more vigilant and forceful than one person could ever be. Our responsibility is to empower all of those watchdogs, and to listen to them, rather than to channel their voice through a single office.
It's not as though the Public Editor was cutting off the flow of criticism, hording it in a single office. You speak of the "modern" world, but you're acting as though the readers' input comes in via paper mail! The Public Editor selected whatever she saw fit to focus on in a column that appeared regularly, but everybody else at the NYT could also see the criticism. If you're just saying that you lazily relied on her and looked to her to answer the complaints, then you are admitting she was not "the reader's representative," she was your representative, saving you the trouble of dealing with the critics. But now you say you want to take on the very role you seem to admit you were avoiding bothering with. Why should we believe that you will take this role seriously if you didn't care about it when you had the Public Editor?
We are dramatically expanding our commenting platform. Currently, we open only 10 percent of our articles to reader comments. Soon, we will open up most of our articles to reader comments. This expansion, made possible by a collaboration with Google, marks a sea change in our ability to serve our readers, to hear from them, and to respond to them.
I've been bothered for a long time that the NYT withholds comments on many articles that need pushback. So now they are going to have more, which is good, but it's still not all, only "most." And it's a bit deceptive to say that the expansion is made possible by Google. All Google is providing is the ability to limit the comments through some kind of automated process. They could have had comments on more or all of the articles, but they chose where to have them, and they apparently still plan to do that, even as they open up more articles to comments, which they're doing because they've built in more power to filter what people try to put up.
We will work hard to curate and respond to the thousands of daily comments....
But what will you work hard to "curate"? What is the standard of what you're keeping out? You won't be revealing that. Maybe you won't even see it. It's the robot's job.
... but comments will form just one bridge between The Times and our audience. We also, of course, engage with readers around the globe on social media, where we have tens of millions of followers. 
We'll see what that means. I'll do my part in social media. I do as much as I can. Who does more?!

We're given the names of 2 editors who will have a special responsibility. Phil Corbett leads a team that "listens and responds to reader concerns and investigates requests for corrections." And  Hanna Ingber will work "to make our report ever more transparent and our journalists more responsive." This seems to be internal work that is not itself transparent. there's no more Public Editor column pointing out problems like lack of transparency, but Corbett and Ingber might make the paper "ever more transparent." How will we know?

43 comments:

Kevin said...

Fake news eventually requires fake comments.

Todd said...

Simplified NTY statements: We will do more to appear responsive to our readers while insuring [through automation] that only right-think comments get through. You are welcome, NTY echo chamber!

Sam L. said...

Pinch is dreaming, and lying to us.

David Begley said...

What a joke. More than ever the NYT is a bi-coastal echo chamber.

Sebastian said...

"How will we know?" I rest easy knowing Althouse is on the case, every single g**n day. Anyway, not to inflate your healthy self-regard any further, this post, like a few other NYT fiskings, is better than anything ever written by any of the "public editors," not least because it comfortably confirms my confirmation bias. Yes, the NYT really is staffed by clueless, biased, dishonest, self-serving prog a**holes, doing their clueless, biased, dishonest, self-serving thing. Mea culpa, Karl Popper: forgive me.

Rob said...

The NYT's abandonment of even the semblance of objective journalism is all too transparent. Having identified Trump as an existential threat, they've decided it's their duty to undermine him and, with any luck, force him out of office. For all the caterwauling about Trump violating norms, it may well be the NYT's departure from journalistic norms that does the more lasting damage to the country.

David Begley said...

" that has been trained to find and flag trolling, hate speech and gratuitous shitposts in the paper's online comments sections. "

And those "shitpists" would be anything that the NYT disagrees with such as being against the Paris Climate Treaty. I'm a proud climate change denier. Lock me up!

David said...

How stupid do they think we are?

Answer: They think we are really really stupid. Beyond Trump stupid. Deplorably stupid.

Bay Area Guy said...

Step in the right direction!

The NYT should now fire all the contributors to the Editorial Page, and replace them with either an AI opinion generator, or, a room full of monkeys randomly typing on keyboards.

It'd have the same intellectual firepower.

Ralph L said...

With thousands of comments, how does any voice of reason get through to the editors, regardless of filtering. The public editor was supposed to be dog-watching for us.

Luke Lea said...

They don't call him Pinch for nothing. The NYT originally made its reputation because all of the other newspapers in New York were unabashedly partisan. Were he alive, Adolph Ochs would be rolling in his grave.

David Begley said...

Ann:

Get over it. You don't count to the NYT. You don't even exist (other than for your subscription). You live in flyover country in a state full of deplorables.

Martin said...

Time for someone to do another "Downfall" dub, of Pinch as Hitler ranting and raving about how useless the Public Editor is and how they can use AI to filter comments, and thereby save their business...

wholelottasplainin said...

Luke Lea said...
They don't call him Pinch for nothing.

***************

Behind his back, they call him Loaf.

Sebastian said...

@BAG: "The NYT should now fire all the contributors to the Editorial Page, and replace them with either an AI opinion generator." Wait, are you implying that writing the NYT EP involves intelligence?

The Cracker Emcee Refulgent said...

Makes perfect sense. The NYT's is selling comfort food, not news. It's working for MSNBC, CNN, and SNL, so why shouldn't they get on the gravy train?

Leigh said...

Abandon all hope, NYT. No one believes anything you say any more. Change your motto to "All the news that's fit to rinse."

rehajm said...

More evidence NYT are propagandists, believe they are responsible for political influence and fear they are losing it.

Unknown said...

"Still a man hears what he wants to hear
And disregards the rest..."

Good job, Pinch. Tune that filter to exclude anything remotely disquieting. And as you go along, and become ever more sensitive to slights and insults, you can keep tuning for ever-more-congenial "facts".

Because that's how media works!

JimR said...

https://twitter.com/RichardGrenell/status/869968104222113792

Nonapod said...

A clever troll should be able to come up with strategies to defeat the AI. Instead of using a term like "fake news", call it "fraudulent infotainment"

Bay Area Guy said...

@Nonapod,

"Fake News"

"Fraudulent infotainment"

"Ersatz Epistemology"

Joe Biden, America's Putin said...

the collective is watching... and washing.

chuck said...

> Fake news eventually requires fake comments.

Comments will be a bridge too far.

Yancey Ward said...

It doesn't take a genius to figure out exactly what will be determined to be trolling and what isn't in the comments sections.

Peter said...

An obvious way to implement this "AI" would be to use the software to do a pre-filter of comments containing objectionable words (presumably with enough intelligence to detect the obvious work-arounds) and then pay an army of politically reliable "mechanical turks" (or volunteers?) to filter the remaining comments.

HoodlumDoodlum said...

We also, of course, engage with readers around the globe on social media, where we have tens of millions of followers.

Yeah, and how many non-Leftists do their writers have blocked on Twitter, I wonder? I don't care if they want to cocoon themselves in their Leftist world--it's their paper and they can do as they like.
It's a bit much, though, to be told I should PRAISE them for these kinds of moves. If they're too afraid of negative feedback to retain someone on staff to "keep them honest," fine, but they ought to at least admit that's what they're doing.
I mean, they ought to if their own credibility w/non Leftists mattered to them...but I guess we already know the answer to that one.

Kevin said...

> Comments will be a bridge too far.

Not for the people who revel in the fakeness. They will cheer the enhanced reading experience.

rehajm said...

The NYT's is selling comfort food, not news.

Yes. At this point how many more subscribers could they lose over this policy?

Bay Area Guy said...

If i may be so bold as to paraphrase a certain famous Secretary of State, WHAT DIFFERENCE, AT THIS POINT, DOES IT MAKE?

Fen said...

NOV 2020 (New York Times Stsff)

We feel an obligation to apologize to our readers. Obviously, we misread the electorate when we forcasted Chelsea Clinton was set to sweep all 50 states.

We don't yet know how we could have gotten this election so very wrong, but if we ever find a conservative, we will be sure to ask them. From a safe distance, of course.

We have a responsibility to our readers to be honest information brokers, so that you can make sound decisions based on the most current and accurate reporting.

With that in mind, we have set the building on fire. We're so sorry to have failed you yet again. Please don't hang us from lamp posts.

Look! There go the Russians! They did this to you! Go catch them and impeachment will be just around the corner.

J2 said...

They already do that annoying thing of awarding gold stars to their most favorite "The Times Picks" to comments. It saves you time getting your thoughts properly aligned. I'd love to know how those comments are "curated".

tcrosse said...

I must protest the recent expansion of the term 'curated'. At one time it applied only to Art and galleries. Now one reads of a delicatessen having a curated selection of braunschweiger, suitable for pairings with curated pumpernickel. Enough, already !

The Godfather said...

It's a shame that the NYT isn't a reliable source of straight, unbiased news, as it once was (many long years ago). We can each pick a news source that will give us the slant on the news that suits us, but when my source and your source disagree about the FACTS, what do we do? Time was, you could go to the news pages of the Times and find the facts. No longer. And there's no source I know of nowadays that provides a reasonably reliable account of the facts. We are living in a post-modern world where there is no reality, only opinions.

RMc said...

If it wasn't for artificial intelligence, they'd have none at all.

Darrell said...

Artificial stupidity when the Times does it.

Kevin said...

I can see why she had to go.

She noted, for instance, in an elementary insight, that a false statement, even one uttered by a person you hate, even a president you hate, is not automatically a “lie.” If you don’t know your false statement is false, it isn’t a lie, and journalists aren’t mind-readers. In a column entitled “Why Readers See the Times as Liberal,” she noted that many a liberal and centrist acolyte of the Times told her that they were seeking other outlets for balance. “A paper whose journalism appeals to only half the country has a dangerously severed public mission,” she said. That such a statement is now considered “controversial” does not reflect well on the media.

JAORE said...

At least the AI can save money. I'm sure the Obama IRS folk would give them copy of the screening software that pointed out the organizations needing extra special scrutiny.

David said...

"I must protest the recent expansion of the term 'curated'"

Those who can, do. Those who can't, curate.

vanderleun said...

It's nothing but "The New York Turds" from now on.

Anonymous said...

I hear ErdoÄŸan and Xi Jinping are interested in adopting the Times's AI.

Anonymous said...

Good day to everyone out there and I am here to share a very great testimony of how I became free and happy today after all I went through with my Ex wife and her lawyer as they tried to end my life after my wife and her so called attorney tried to ruined my life. I was with my Ex wife for over 10years and we both have worked together that we had a joint account both in Europe and America because we loved each other so much that I never expected what happened to me later. My wife filed for a divorce and tried to claim all the money we worked hard for both in Europe and America including our only child Mira. So I almost gave up my life if not for the help of my very good friend who resides in Europe where I worked for over 7years and we have been good coworkers for years that introduced me to a hacker whom I contacted to help me hack into the accounts and move the money for me not to end up committing suicide at the age of 56 years. At first I was scared that if he falls and get caught what’s going to happen but my coworker assured me that his a genius and very calm and ready to listen to my problems so I proceeded with him to handle the hacking of the bank accounts to move some of the money for me without anyone getting to know about it till date. He was able to move %80 of the funds both in Europe and America via online banking system that I am now leaving in Australia trying to start up a new life and family too. Please I know too well there are lots of people going through this out there and I pray for you to be strong until you get to see this post that will change your life in less than a week of working with this amazing hacker. My name is Patrick Fisher and I tell you that his the best and very good hearted man to help you and I don’t know if he does any other hacking job but I can only testify for what a did for me. May God bless each and every one of you going through hard times in their homes today? be strong and I hope everyone tells their stories someday.
Email: michael.l@hackermail.com
Telephone +1 646-652-6107

Tim Cunningham said...

For those on the left, no big deal.
For those on the right, additional proof of progressive media stupidity.
For those who love the press as it should be, seeking and publishing truth wherever it leads without fear or favour to one's favourite politicians or policies, this is regrettable. Censoring feedback like this will not help the Times thrive.