App Link: https://apps.apple.com/us/app/boring-report-news-by-ai/id644...
He said that today, but in an interview with Bloomberg. The source article[2] just illustrates it with an archive photo from 2022, when he testified in a Senate hearing. Similarly, the Disney article[3] starts non-sensically The Disney+ logo was displayed on a TV screen in Paris on December 26, 2019. Disney shares decreased by 9%[...] (I don't think displaying the logo 4 years ago is to blame).
I suppose you should just stop parsing image subtitles. The two articles I checked were otherwise accurate.
[1] https://www.boringreport.org/app/all/645cfc85bab323b21e6195e... I had to use the developer tools to copy paste the text, obnoxious. You also can't right- or middle-click the source link (to copy it or open it in a background tab). Don't hijack basic browser functionality.
[2] https://www.cnbc.com/2023/05/11/jpms-jamie-dimon-warns-of-ma...
[3] https://www.boringreport.org/app/all/645d0cebbaef7c040f89ca4...
Start with those and then figure out how to scrape a site as your input and spit out the existing API format and you'll come in through a clever side route, essentially having a two phase assembly line.
Also this will allow users to customize their "feed" as a free side effect of the architecture and furthermore you'll be able to isolate your scraping -> API transform on a per site basis, also as a free consequence and lastly, you can parallelize the work much cleaner and even have the public add their own "transformer" for their favorite news site
I mean, it's great because it's accurate. Half of the "news" we're fed is sensationalized so we'll click on it and it's really nothing but it gets us riled up about something that is effectively meaningless to us. This just brings reality to the forefront and makes me realize I don't care about the news lol
Thank you though, this is awesome!
This is sort of like how you recalibrate your tongue to a low-salt diet if you stop eating salty food for a few weeks.
Maybe the pre-clickbait era version of you would have found more of these articles interesting? So maybe it's not that you don't care about the news, but your "base level" of what's interesting is different due to being used to clickbait.
That's a feature, not a bug. The 'news' is 90% useless, if not descrutive, information. Ask yourself: what percentage of the news is optimal to maintain your worldly wisdom? I'd guess about 1%. So getting it down to 10% is half way there, on a logarithmic scale.
---
To your request of my opinion of the manner in which a newspaper should be conducted so as to be most useful, I should answer ‘by restraining it to true facts & sound principles only.’ yet I fear such a paper would find few subscribers. it is a melancholy truth that a suppression of the press could not more compleatly deprive the nation of it’s benefits, than is done by it’s abandoned prostitution to falsehood.
Nothing can now be believed which is seen in a newspaper. truth itself becomes suspicious by being put into that polluted vehicle. the real extent of this state of misinformation is known only to those who are in situations to confront facts within their knolege with the lies of the day. I really look with commiseration over the great body of my fellow citizens, who, reading newspapers, live & die in the belief that they have known something of what has been passing in the world in their time: whereas the accounts they have read in newspapers are just as true a history of any other period of the world as of the present, except that the real names of the day are affixed to their fables.
General facts may indeed be collected from them, such as that Europe is now at war, that Bonaparte has been a successful warrior, that he has subjected a great portion of Europe to his will &c &c. but no details can be relied on. I will add that the man who never looks into a newspaper is better informed than he who reads them; inasmuch as he who knows nothing is nearer to truth than he whose mind is filled with falsehoods & errors. he who reads nothing will still learn the great facts, and the details are all false.
Thomas Jefferson, 1807
---
The most interesting thing is that this would have sounded quite hyperbolic but 10 years ago, and probably near to completely unreasonable 30 years ago. Yet now? It sounds completely reasonable. Like so many things in history show, the era we're entering into is not some wild uncharted domain, as it sometimes feels. Rather the era we all lived and grew up in was the weird one. Now we're simply returning to 'normalcy.'
[1] - https://www.founders.archives.gov/documents/Jefferson/99-01-...
Half? It's higher than that.
The other issue is - and your quoted "news" hints at it - a significant amount of what we're fed isn't news at all. Yes, it's something that happened. That doesn't make it news.
So while this app might remove the hyperbole what about the things and details that are being ignored? How do we fill the gaps so we're getting a more complete context?
I tried doing the same thing as OP with chatgpt and came to the same realization so I stopped.
Everyone wants to consume their work, but they need people to pay for it so that they can keep working, but people won’t consume their work unless it’s sensational.
I think this is only true because we allow unlimited sensationalism. It's just like any other junk intake (candy, social media, gamification, etc) - it's easy, it has an immediate reward, and it's not valuable. But out bodies and brains were conditioned by evolution to only care about the first 2, so junk news is just hacking our brains (or exploiting our vulnerabilities).
To be blunt, we've turned something that should serve us into a parasite. We need to have awareness of this and put create standards and/or regulations. When the playing field is level and (almost) nobody is writing clickbait, our brains will return to normal.
Previously translation has focused on going between languages.
I think there's just as much benefit to translating within a given language.
* Translate corporate speak to plain English.
* Translate passive aggressive to calm and peaceful.
* Translate sensationalist to neutral (like the OP).
* Translate implicit and heavy with subtext to direct and assertive.
Of course the utility will depend on the training data you collect. For example to go from corporate speak -> english you would need thousands of pairs of "translated" sentences. Or you could use an LLM to translate for you, paying as you go
One time, local-dialect subtitles completely changed a movie experience for me
It was an American movie. Where I was, they usually watch movies with english audio and neutral Spanish subtitles
This time, instead of neutral Spanish, they used local Spanish with slang… wow, what a difference in the way the movie felt
Having local language and slang can convey meaning and emotions so much more effectively
Otherwise, it's someone typing as fast as the movie goes on the first viewing.
> The Legend of Zelda: Tears of the Kingdom Performance on Switch
Article's title:
> Zelda: Tears Of The Kingdom’s Performance On Switch Sounds Like A Minor Miracle
I find the Boring Report summary to be much more informative that the article, it's like they trimmed away all the fat.
The fact that the game runs at 20-30 FPS doesn't seem like a "minor miracle", by 2023 standards that's barely acceptable. And I understand the limitations of the Switch hardware, I'm not trying to insult Nintendo or the Zelda games, but I have hard time using the word "miracle" to describe the performance of a game that runs at 20-30 FPS.
For the record, I own a Switch and I'm thinking about buying the new Zelda game... I just like facts more than spin.
"[Zelda] Performance on the Switch is Passable and Impressive under the Limitations of Hardware"
> The Legend of Zelda: Tears of the Kingdom performance: 20 to 30 FPS on Nintendo Switch
That would generally make it unnecessary to read the article.
Edit: after reading some of the content i really like it!
In 1922, a man named Nick Carraway moved to West Egg, Long Island, a region populated by newly rich individuals who have established social connections. Nick has taken residence in a small house next door to a mansion owned by a man named Jay Gatsby.
One big complaint I have for reading news through RSS, is that there's no natural hierarchy/priority to the news. There's no front page, no headline, no size in RSS feeds. Given the way news agencies generates those feeds, there are _tons_ of repetition, tiny updates, some insignificant one-liner interview about some significant events. Not to mention the "no update at this point" updates. Entries that are not informative look exactly the same as---but often outnumbers---the entries that are informative.
An ideal news feed processor to me, would be one that reads through last weeks RSS feeds, and merges the all those tiny updates into coherent articles, ranked by the significance of the event. Sort of turning newspaper into a journal.
The merging and reflow should be well-within an LM's capability. However, I'm not sure if OpenAI's API can swallow an entire week's worth of RSS, or produce multiple full-sized articles, but this is something that I'd like to try when I get some free weekends.
“hey, remember how everyone was panicking about the price of eggs a few months ago? Well, prices are normal now but only one person wrote about it so you probably didn’t hear that”
The main differences compared to what you are thinking are two things. One for the `Significance of the event` I've used the number of publishers talking about that event. So more publishers == more important. Two, I've done this in a daily fashion instead of a weekly report.
I can also confirm that the LM has the capability to do at least a days worth (2500+) of articles. I would doubt its capability to produce an entire article but it does a great job at a small summary.
Here is the link if you wanted to check it out. https://apps.apple.com/us/app/quill-news-digest/id1669557131
Maybe the solution isn't to pretend things are just fine. People keep trying to paint this as some sort of "sensationalism" and "I want just the facts" but the facts are that thousands of people die every day from cheaply and easily preventable illnesses and issues while people who can literally self fund rocketry get to accumulate even more wealth and power.
There's a difference between "remember the maine" and "hey women right now are literally dying because they can't get abortions to remove dead tissue inside their body because of some completely different person believes their religion says doing so is a crime"
Being angry from news like this isn't a bad thing. You SHOULD be angry.
Why?
What are you going to do?
Just get high off the dopamine drip fed by social media rage cycles? Angry rants on Twitter? Maybe start a Youtube channel and make a buck or two with banal political essays?
Recognize that the news is designed to make you angry because that anger prevents you from taking any form of effective action, it misdirects your energy and consumes your focus. It is a means of control, leading you to confuse catharsis with praxis.
By all means, be concerned about the world, injustice, social inequality, etc. There is a lot to be concerned about. But anger - much less the indignant virtue signaling anger you're displaying here, is a waste of time.
This was a key step towards improving my own mental health by unplugging from news. I do care about things like abortion and crime, and I vote accordingly, but the amount I am outraged does not affect the outcome. So consuming media that merely outrages me and tries to grab my attention, but not in a way that galvanizes me to actually do something about it beyond voting and word of mouth, just worsens my mental state to no benefit.
I'm not sure if they always achieved that, but it was at least a goal for readers to be able to skim feeds to understand what was going as quickly as possible. Such a policy was feasible because the readers were paying customers as opposed to web news which are often funded by advertisers or worse.
The prompt asks to rewrite the article in an unbiased way and to also expose biases in the article. It feels 'good' to use.
Demo here: https://twitter.com/vladquant/status/1647042056139968512
Code and prompt here: https://github.com/OrionBrowser/ProgrammableButtons
A reasonable premise! But easier said than done. I wonder how this app counteracts the hallucination and lying behavior of LLMs.* Would be pretty bad to trade off easier-to-decipher human bias and sensationalism for distorted truths and lies from an obfuscated sequence of dot products!
* I assume they are using LLMs because they state:
> By utilizing the power of advanced AI language models capable of generating human-like text,
I don't think there's a generalized solution to this problem for all information domains, so when search engine companies implement it it'll be low quality. What remains to be seen is if the money is in being a curator/aggregator within a niche, as Boring Report aims to be; or if it will be in selling the specialized summarization tech to the content creators directly - for use by them when they publish. I think the latter leans B2B and will have higher quality since the content creator signs off on it. But we'll see. Either way, the right mental model for LLMs may be to treat them as memetic compression algorithms.
I genuinely think there is a huge underserved market for a "world's most boring news and weather site". Almost everybody I talk to on all sides of the aisle recognizes that clickbait news is one root cause of lots of problems and want an alternative. In fact, in some ways, Hacker News is that site for me.
That said, I don't get why an app not a mobile-responsive website.
When considering installing any app, your first question should be "is there any benefit to the user of making this an app instead of a web page?" The answer is very frequently no, which means effort spent building the app and getting me to install it must be for a purpose that's not in my interests, such as tracking.
It's a hard problem! But I'd guess that not many people will stick with this as their news source, because it won't hold their interest, because it doesn't include all that information about why they might want to care.
https://chrome.google.com/webstore/detail/clickbait-remover-...
Further discussion on HN:
https://www.google.com/search?q=site%3Anews.ycombinator.com+...
this probably functions decently well as a "europe vs america" discriminator.
PBS News Hour also does a great job of this. Just a few global and national stories a day, but a stark contrast from the emotional triggers from other news outlets.
Unfortunately have to delete the app for now. Will check back later to see if they’ve figured it out.
"Former President Donald Trump appears on CNN town hall forum in what is considered the first major television event of the 2024 presidential campaign (More) | See all 2024 presidential candidates so far (More)"
(More) is a link to further information.
Top story for that day (today) was Title 42 Ends and second was Santos Charges Unsealed
Media literacy is about way more than about the wording of headlines. It's also about understanding why a headline was selected, who benefits from a story, whether the story is internally logically consistent, and why were the people quoted selected, context of the story that you wouldn't know just from reading it, etc.
I say this as someone that wrote a browser plugin to do something similar in like 2011 by screening words that indicated the headline was pointless.
It seems like the latter might be better, since it ensures the headline actually matches the article below, as opposed to relying on a likely-clickbaity title that would no longer match the desensationalized article body.
Probably worth testing both ways to see what the results look like!
(also, that click jacking to prevent copy/paste is a PITA)
Only the passing ones are then delivered to me.
I'm not releasing this as a product, it's too simple, but it works surprisingly well, and it's trivial to add criteria for what you deem to be important, or not.
Here are some examples:
Musk Legal Issues: https://markets.sh/stories/elon-musk-s-legal-issues-2023-05-...
TikTok Ban: https://markets.sh/stories/tiktok-ban-2023-05-11-01-16-03
Bard: https://markets.sh/stories/google-bard-ai-chatbot-2023-05-11...
With the current advancements, is there finally a browser extension that just hides clickbait titles/thumbnails?
Full thread: https://news.ycombinator.com/item?id=35795388
Edit: To give an example: Coming across the headline "Popular action series is cancelled after just one season", the summary really provides everything I want to know (or at least: to let me decide if I want to spend time actually reading the article):
> Cancelled after one season: CBS has axed action series True Lies, based on the 1994 film, after struggling to find its own identity.
> Mixed reviews and lack of audience: The show failed to gain a large enough audience despite starring actors like Steve Howey and Beverly D'Angelo.
It's ironic that this tool is essentially reverting news reporting back to what journalism is supposed to be - factually reporting notable events that have already happened. It's bizarre how acceptable it now is for 'news' to include 'opinion and speculation'.
- The world section is super weird especially where important geopolitics are mixed in with "someone dies in a car crash in the UK".
- The web app is pretty bad for what could be much better off as a static text document. Expanding articles is unnecessary slow. Why do I have to wait 3 seconds to get 1 extra sentence and a link? Why abbreviate 3 sentences into half of a sentence at all? The full article also comes up at the bottom of the screen while locking out the rest of the app - trully weird. Honestly, it's not very usable at the current state.
https://www.newsminimalist.com/
Absolute gem, have been using is everyday now
E.g.
- how much are we borrowing this year and the last 50 years
- what is net migration for the last 50 years
- what is inflation for the last 50 years
- what are the crime rates for the last 50 years
- what are the violent crime rates for the last 50 years
Don't know why I picked 50 every time, but I'd just love some key, well-contextualised data that we could all agree on. At the moment it seems as though people can't even agree on basic facts, and anything that helps with that would be awesome.
Well, if they can present the rate of crime underreporting, and how it has changed over time, they're welcome to add it to the graph :)
Stay vigilant.
https://www.boringreport.org/app/all/645a8edbcdaefdcdfe3806b...
From A Random Article[0]:
"In the second case, the court ruled in favor of Louis Ciminelli, a Buffalo developer convicted for taking part in rigging the bid process and property fraud. The justices reversed a lower court ruling based on a theory of law that the government later abandoned."
That might be "boring" to the point of "uninformative." It also gives me strange reminders of the book "Brave New World," and I think most of the authors so edited might feel the same way.
I would presume to _add more_ to the article then to _take away_ from it, in some cases, to the point of blinding the reader from any fact whatsoever.
[0]: https://www.boringreport.org/app/all/645d396cf7c90670355a6c7...
The ambiguity of "metal object" kind of makes it more sensational than "possible meteorite".
News media needs to make money. To do so it needs eyeballs. Eyeballs, despite our best efforts and intensions, can't help but be drawn to sensation (and blood). The news media thus selects for sensation and blood.
The add-on effect of this that makes it all matter: recency bias as well as some other cognitive quirks we all have mean that when the news media selects for sensation and blood, we start to believe the universe is a lot more sensational and bloody than it really is.
This is demonstrable: publish a story about a freak one in a million train accident and people stop riding the train and start taking their cars, despite risk of accident in car being by every way of measuring it drastically higher.
I'm not saying that news media shouldn't talk about train crashes, I'm saying I'm not sure we've figured out yet how to remind people that the world is actually mundane. People watch the news to "be informed," but in actuality their perspection of reality is being warped.
The news would have you believe crime is going up or down based on whether it makes a good story, when (depending on where you live) the relatively minor adjustmens in these statistics really don't affect you, every walk to work statistically speaking will leave you unmolested.
The media would have you believe that there are riots across the country in response to the cops killing an unarmed person, again. When in reality, most protests are really quite boring, just a lot of people walking around and maybe occasionally chanting. If there even is a local protest.
I don't know what it will take to remind people that the world is boring, but right now the news media is motivated against this, because the angrier and more scared we all are, the more we click in.
I think what you're really saying is: "I'm less interested in the work this person did to make something cool; selfishly, I'd like to know if it follows my biases so I can judge it based on politics."
Or you're just curious how it works and haven't spent the effort trying it out, which is valid.
Could it not be that as a person with an interest in AI and general technology I am wondering if there is any detail on what does an "AI" look for to determine a "sensationalist" title and how does it "desensationalize" it?
I have scanned the twitter feed and the app and just wondering if there is anything on how it works. The answer can be no.
NBC News, almost every week: "20 million Americans threatened by severe weather this weekend!"
"Bomb cyclone threatens 10 million people"
"Atmospheric river menacing 18 million people"
Translation: A rainy front is moving through several states.
The atmospheric river post would be an example of the second headline and should still maintain the necessary urgency and implied danger.
What I hate about these particular (nearly verbatim) examples is the summing of the entire population of every state that might be touched by a system and barfing it out as if it's a body count.
And why all of a sudden does every weather phenomenon (or variant thereof) have a new sensational name?
And on a California-centric tangent: I also detest meaningless labels, which CA loves. For example, there's a RED FLAG WARNING! WTF is that supposed to mean? Are millions of people threatened by red flags? OH NOES!
And finally there's "sigalert." Insert giant eye-roll emoji.
I'm far more interested in what a sensational news org decided to not include.
Question: How do you intend to make money from/monetize this?
The reason I ask is because I have several "pet" projects like this I've been meaning to develop but with the time cost and financial cost it has to be something I can at least cover the server and maintenance costs.
I would also like to see an RSS reader that does the same to my own feeds, but that could be cost-prohibitive.
Those of you who think it isn't... let me guess... You didn't major in sociology, anthropology, or psychology?
Those of you who respect sociology/anthropology/psychology and yet still think that "sensationalism" (presented in quotes because that is a socially-constructed, loaded term) is something that distracts you from real news, I'd like to hear some of your nuanced thoughts on this. Because it seems to me that one man's "sensationalizing" is another man's "helping you understand the relevance of recent events to your life." And that's quite important.
There is something like this in 'A Fire Upon the Deep' by Vernor Vinge; his intergalactic societies translate their alien languages - and incompatible methods of expression - using an application similar to this one.
Very cool whenever sci-fi becomes reality!
I believe Google News does some bundling of topics.
Though I also recognize that topics are hard to define as they could be chosen arbitrarily wide. Everything could be a topic. The ultimate topic would be “The Universe” which contains everything.
Don't rephrase it, just show me all of the sentence fragments that need to be rephrased. Or better still, make it a 'lint' tool that rates articles and websites by the density of manipulative verbiage they use.
These sites will piss you off and teach you nothing in the process.
These sites will piss you off but actually teach you something.
These sites will placate you (I'm looking at you, Ted Talks).
It's shows left, center, and right bias of new articles.
Read the 'From the center' version of articles, and the bias and sensationalism has been removed - https://www.allsides.com/story/media-industry-cnn-staffers-c...
I don’t think we can boil down most things into three separate POV. Not everything needs to be looked at through a political party’s perspective
What other system would you recommend for evaluating political bias in news?
i'd also like to see a service which shows me the news about subjects which have managed to stay in the news for at least a week. just drop all the 24-48 hour rage/hype cycles.
https://thewire.in/media/naj-duj-hindusthan-samachar-prasar-...
One suggestion (maybe you’re already doing this), but it’s not just sensationalism but also the bias towards negative events being reported that could be addressed. “If it bleeds it leads” is a famous adage in the news world. I would love to have control over the share of negative events I see in my feed. Contrary to the popular news, the world is a far healthier and safer place than it was in the past.
You can see the share here: https://www.facebook.com/groups/dullmensclub/permalink/12449...
is well summarized by "The Christie Receives CQC Rating".
(also overriding my mouse buttons to prevent me selecting text? How very passe.)
Wonder if you can make Fox News boring too.
0, https://www.forbes.com/sites/markjoyella/2023/02/01/fox-news...
Question for the engineers working on this: Does this just summarize one article or at a time or cross reference multiple different sources for the same event?
Today, many news articles contain eye-catching titles and content that can lead readers away from the central facts. Boring Report uses OpenAI's language models to take these types of articles and changes them into the content you are reading now, with the goal of aiding readers in paying attention to the primary information.
Have you considered extending it further so that instead of tackling news articles individually, it would try to find and aggregate / juxtapose everything prominent that is published about some particular news item in a similar "boring" manner?
Today: "Brits to bask in 20C sun" (that's 68F) Yesterday: "Yellow thunderstorm warning..set to spark travel chaos"
Just a tad over the top! Genuine weather extremes in the UK are pretty rare :)
Maybe that's an another app idea: remove "both sides" and "horse race" nonsense.
AI is no longer artificial intelligence but algorithms and text generators.
He does not need de-sensationalizing but he badly needs summarising, which seems adjacent. That Matt Levine refuses to use the services of an editor is one of the greatest tragedies of our era.
The next step is to give each outlet a sensationalism score and then normalize their headlines according to how they usually sensationalize.
This needs work to be what you claim. You've baked in bias and report it without emotion
https://www.boringreport.org/app/all/645c525a001a7a3765dbc4c...
Or are you complaining that it's there at all?
They could also use more right leaning site sources in general. It looks like they mostly have the token Fox News source and that’s it. It would be much more useful if I could either add my own sources, or for every left leaning source a right leaning one was added.
"just hand it to OpenAI and :shrug:" is not a responsible answer.
Yeah, AI can totally decide what's real and what's not. Sounds great
That doesn’t tell me what he was claiming or how it was shown to be false. Whether or not someone will trust that conclusion will just depend on whether they tend to trust CNN, or their pre-existing bias on Trump/Jan 6/Pelosi.
Reading the original article his claim was that Pelosi was in charge of security as speaker of the house, and that is false because that’s the responsibility of the Capitol Police Board. I think that’d be a more useful summary of that section and a human wouldn’t have stripped away the specifics quite so much.
* rapturous applause * if this works.
