Show HN: A new way to fight manipulation in news | Hacker News Reader

Show HN: A new way to fight manipulation in news

dailyedit.com

15 pointsstevepeg9 comments

Forgive my skepticism, but I've seen this before. The Knife of Aristotle[1] made similar claims and actually seemed to be doing a good job, until it was discovered to be a front for a sex cult[2]. One of their primary methods of highlighting bias in news media was to identify unnecessary adjectives and loaded terminology, which is IMO still a valid concept to paint a picture not necessarily wrong but intentionally misleading slant (e.g. "The deadly insurrection on the U.S. Capitol" is still an oft-used phrase). I am also not convinced by vague "we use machine learning" marketing terms without some very specific details about the training data, methodology, algorithms, fitting, human curation, etc.

How does The Daily Edit differ, and what assurances does a user have that it's not a similar front for covering up stories your staff doesn't want to be propagated?

[1]: https://mediabiasfactcheck.com/the-knife/ [2]: https://artvoice.com/2018/08/06/the-knife-media-has-ceased-p...

Skepticism welcomed, it's a crazy space so the more of that the better!

We differ by giving the reader the tools to form a better opinion for themselves rather than telling them what's wrong and making ourselves some oracle or source of truth. We think that's where so many of these attempts go wrong.

Our highlights show related details that were found in other articles, but were not covered in the article being read. We show all of the sources of these so the user never has to take our word for it, they can go there and see. This part uses an ML model trained on the MNLI and SNLI datasets. I'm happy to share all the details of the graph and training method privately if you send us an email.

We do also highlight some passages of text for being potentially misleading but this uses no ML whatsoever. We suggest missing sources of data ("according to our sources"), missing reference ("a recent study suggests" - with no reference or link), and scare quotes. We make sure that the item is verifiable by the text the user is actually looking at. We do not attempt to cover subjective items like hyperbole, slippery slope fallacy, etc. Each of these items is presented as a suggestion intended to make the reader pause for a second.

Your last question is hard to answer since I'm not sure what assurances anyone could give apart from "we promise", I'll do my best though. We're a startup of 6 people who all care deeply about the quality of news. If we fail to cover an article you should call us out on it and find out why then share that with everyone.

Right now there are a number of technical reasons why that might happen such as the insane mess of HTML some publishers use, it can be hard to parse.

Hi everyone, I’m Steve, one of the team working on The Daily Edit.

We’ve built a news aggregator which shows you every detail that has been reported on a story regardless of which article, or which publisher, you choose to visit. Readers can choose whichever angle they like and they’ll still see everything that’s been reported, from all sides. They don’t have to visit several articles to do this, it all happens in one place.

We’ve put some examples of what it all does on Twitter: @dailyeditapp. If you find more great examples please share them with us!

How did we get here?

Myself, Craig and Mich work(ed) as pilots for an airline out of Hong Kong. Over our careers we’ve seen that all of the current issues with the news happen everywhere around the world and, like everyone, we’re tired of it. During COVID our flying rosters were reduced significantly so we had a bit more free time than usual. We used this time to have a go at building something that can help fix this mess.

Initially we made a tool that highlights questionable pieces of text like logical fallacies and such. We tried this out with some friends and kept getting the same feedback. Nobody really cared that much about the text, they all said the real manipulation happens through omission.

For a while we discussed how to do this without falling into the fact-checker trap. Fact-checkers are great but every time they show an error a new group emerges from the mist to spend the rest of their lives arguing against it.

Eventually we realised that if we could aggregate deeper than just the articles, go right into the text itself and pull out details, then we could make this work. We got a small proof-of-concept up and running, put together a team and got to work building the real thing.

We hope you like it! It’s early and there will be bugs, but we’ll squash those! We have a lot more stuff coming over the next few months including a web app and browser extension. If there are particular things you think would be awesome to have, we’d love to hear it!.

If you have any feedback at all feel free to email us at feedback !at! dailyedit.com

I suppose the true problem is to understand the source of financing behind the media organizations. The ideas are pretty easy to understand, what drives them. But in this case you gotta follow the money.

I took a current events class in high school that challenged us to read three articles for every topic. One from the left, one from the right, and one on the fence.

This exists as a website today:

https://www.allsides.com/unbiased-balanced-news

While I'm not sure of the value prop from your app today, perhaps there are ways to help inform the user regarding what your "trust index" means to them.

I think one of your cool features is the "duplication detection" where you find how many times a passage or article is duplicated on the internet. I would find that especially useful to the "trust index" if one side had significantly more articles duplicated than the other.

It's good to hear that media literacy is being taught like that in high schools. Education on the topic is the best sure-fire solution in the long term.

We really like AllSides but where I think our product differs is that we say that sides don't really matter if you get to see everything reported on the story from all publishers.

If you go to an article from the right in our app, we'll inject content from the left (and everyone else) in-context when you tap on a highlighted piece of text. And vice-versa, if you go to an article from the left, you'll also see reporting from the right (and everyone else). You don't have to go looking for the other articles.

A clever news reader will go out and read several articles from different viewpoints on a particular story, but the idea here is that it shouldn't be a manual process when technology can present it to you conveniently.

Cool idea regarding duplicate detection, but it would have to be from a neutral perspective and not point at the left/right as that will only widen divides further. Maybe a list showing all the details that appeared in a story ordered by the number of articles that covered each.

is there no web version?

It's almost ready! It works together with a browser extension which adds the highlights and extra information to articles so we're in the process of tweaking that interaction.

We decided to release the mobile app first since almost 70% of people reading news online do it on their phones.

ah. i only check the news on my computer.

good luck!

Reply on news.ycombinator.com