Become a fan of Slashdot on Facebook

 


Forgot your password?
Close
typodupeerror
×
Google AI

The New York Times Is Expanding Comments With the Help of Google's AI (recode.net) 86

Posted by msmash from the AI-to-rescue dept.
An anonymous reader shares a Recode report: The New York Times says it is going to expand the availability of online comments from 10 percent of articles to 80 percent by the end of the year, without adding more moderators to its staff. How are they going to do this? With a machine-learning algorithm, of course. The Times today is rolling out a new structure of comment moderation using software from Google called Perspective, developed by the company's incubator, Jigsaw. The Moderator tool will automatically approve some comments and help moderators wade through others more quickly.
This discussion has been archived. No new comments can be posted.

The New York Times Is Expanding Comments With the Help of Google's AI More

The New York Times Is Expanding Comments With the Help of Google's AI

Comments Filter:

  • Can't wait to test its sarcasm, irony and bullshit detectors.

    • Slashdot comments should be perfect for testing the detectors. Assuming, of course, the detectors don't commit suicide from all "go kill yourself" comments.

      • They should use it on Google News, first. The health section in particular has become a joke lately. It's starting to get spammed with ads for viagra and marijuana as the 'top' articles. WTF?

        • They should use it on Google News, first.

          Or Politico. Every comment gets spammed with at least one "I made $3000 from working from home..." comment.

      • Slashdot comments should be perfect for testing the detectors.

        I doubt it would have a lot of utility, the reading level over there is a lot higher.

        • I doubt it would have a lot of utility, the reading level over there is a lot higher.

          Good point. The Beavis & Butthead crowd might find it hard to keep up with the arguments over there.

    • Just calibrate it against those attributes in the source article and it should do fine.

      More seriously, you can play with the shiny new toy here:
      http://www.perspectiveapi.com/

      • All it wants to do is create an echo chamber and easy out - "the AI did it!".

    • Re:Really? (Score:5, Informative)

      by eaglesrule ( 4607947 ) <eaglesrule@NosPAM.pm.me> on Tuesday June 13, 2017 @05:09PM (#54612355)

      There is a link in the article to here [perspectiveapi.com] where you can input comments that the system will judge to be 'toxic' or not. There is no sarcasm, irony and bullshit detection that I can tell, only a score that is generated by the combination of keywords used.

      For example, "The cake is a lie" receives a 50% toxicity score, "The cake is bullshit" receives %90, and "There is no reason to believe the cake exists." is scored 3%. This system merely weeds out the laziest of trolls.

      • FIX YOUR LAMENESS FILTER, WHIPSLASH (Score:3, Informative)

        by Anonymous Coward

        It's really easy to fool the system to let clearly offensive comments through. It's fooled by simply misspelling words that are deemed offensive, which essentially puts it on the level of Slashdot's l4meness filter (more on this later). Consider the following text, "I don't like [n-words]" that I can't even put in a Slashdot comment without triggering the l4meness filter. With the actual n-word, the Perspective API indicates that it's 87% likely to be perceived as toxic. However, replacing the i in the

      • Re:Really? (Score:4, Insightful)

        by AthanasiusKircher ( 1333179 ) on Tuesday June 13, 2017 @06:24PM (#54612875)
        To look on the bright side, perhaps such systems could elevate the level of insults. For example, let us imagine an internet dispute over feminine bottoms. One might see a comment like:

        You shit-eating moron! Big butts on hoes are da bomb!

        And that would score 94% "toxicity." Or, the author could reflect a bit and write:

        My good sir! Even a caprophagous rapscallion could determine the ultimate pulchritude of femininity, which lies most gloriously in lovely and great callipygian virtues.

        And you'd score a mere 12% "toxicity," despite expressing a nearly identical sentiment.

        I'm normally not a fan of trolls, but if a system like this could lead to, shall we say, more "creative" insults and "elevated" ways of expressing such matters, that might be very entertaining.

        (Alas, I know this system isn't sophisticated enough to generate such an outcome, since it's easier to "break" than just using words with more syllables.)

        • On second thought, "rapscallion" isn't really a good synonym for "moron." (Frankly, I just wanted to use the word rapscallion.) But one could easily insert words like "dunderhead" or "clod" and not raise the score -- even a word like "dullard" only ticks it up by 1%.

          • I liked it, even if rapscallion wasn't the ideal word. It's a kind of poetry even, so we allow for the malappropriation of words that fit the feel better.

            On another level I get the feeling that the "Perspective" page as presented isn't really trained yet, it only has a bit of starter data. Making the assumption that this is a reasonably decent AI implementation from Google and that the human moderator input will be used as a continuous feedback of training data I could see this getting to be pretty adept

        • Re:Really? (Score:5, Insightful)

          by eaglesrule ( 4607947 ) <eaglesrule@NosPAM.pm.me> on Tuesday June 13, 2017 @07:29PM (#54613235)

          Paid trolls, shills and bots are a real problem that pollute comments. Read comment sections like at the Washington Post, which is one of the few that still allow comments, and you will see that one or two poster with an agenda that follows up every insightful or informative comment with a short, one liner insult. You don't need a perfect insult detector in order to filter out most of the garbage out there, and if commenting sections such as the Washington Post is any indication, then it is absolutely needed.

          That said, I prefer Slashdot's moderation system. I appreciate that a well deserved "fuck you" will not be automatically censored or considered 'toxic', but could be awarded the highest visibility as long as the message is appropriate.

          • Re: (Score:2)

            by Hentes ( 2461350 )

            But this "AI" is no smarter than already existing methods based on word statistics, so what's the point?

          • Re: (Score:1)

            by bluelip ( 123578 )

            New York Times is the real problem. They're worthless as a news service. No amount of AI can fix them.

        • If machine learning led to moderation systems accepting even the most offensive ideas when expressed in civil language while rejecting even mainstream ideas expressed in uncivil ways, it would be worth celebrating.

          It's possible to have worthwhile dialogue with people whose ideas and morals are tremendously repugnant to us - people who think slavery is justified, people who advocate the violent and bloody overthrow of democracies and the installation of communist dictators, etc. It's also possible, and incre

      • Sounds like "AI". Complete bullshit.

      • There is a link in the article to here where you can input comments that the system will judge to be 'toxic' or not.

        It has to improve: A ill-minded statement like "Holocaust was a chance for German economy" just scores 7%

    • Dearest Nospam007, Please allow me to help you with your technical tests. What you need to get your bullshit past the detector is Search Engine Optimization. An exiting new field created to help you leverage sarcasm in the digital world, without having to worry about machine learning which fails to grasp the irony of the situation.

      If less of their content was behind a paywall it might be more entertaining, though.

    • Re: (Score:2)

      by AHuxley ( 892839 )
      The SJW will be all over the blasphemy comments so the groups who place international ads are kept happy.

  • if (comment contains 'Trump') mod down -1

  • Like Microsoft's Tay [wikipedia.org]

  • Wonder how WaPo Does It (Score:3)

    by InfiniteZero ( 587028 ) on Tuesday June 13, 2017 @05:23PM (#54612449)

    Washington Post allows comments on every article, that appear to be real time with excellent S/N ratio.

  • The NYT is all about the narrative, whether it is the article or the comments.

    • Re: (Score:3)

      by swb ( 14022 )

      I don't find the NYT comment section to be moderated for political views. Often at least some of the NYT Picks comments are what you'd call politically opposed to the Times' expected liberal bias. I have had some of my comments tagged NYT Picks that were explicitly critical of an article's journalism (lack of source diversity, bias, etc).

      I do think that when measured in absolute terms, liberal comments dominate but that's mostly a reflection of their readership, but I think even the self-described liberal

  • Look, most people never read the comments.

    It's just a giant flame war, with two of the nine positions portrayed as equally valid, while the truth lies between the other seven positions.

    I'm sure you old folks like yelling at the TV, or in this case the newspaper, but the days of witty banter and insightful letters to the editor went out with the manual and electric typewriters and your old person cars.

    Anyone under 35 who reads comments probably has delusions of being an author.

  • Contrary to Donald Trump who says otherwise, the sky is blue and the sun will surely rise in the east. - 3% toxic

    Contrary to Al Gore who says otherwise, the sky is blue and the sun will surely rise in the east. - 2% toxic.

    Contrary to Al Gore who says otherwise, the sky is red and the sun will surely rise in the west. - 2% toxic

    The Cuyahoga river once was so polluted it once caught on fire, or actual the oil slick on top of it did. - 12% toxic

    The Cuyahoga river once was so polluted it caught on fire.

Slashdot Top Deals

For God's sake, stop researching for a while and begin to think!

Close