By definition, news is things that are newsworthy. I.e. the most common things to tech receive coverage are the uncommon to things.
If four people get shot over the weekend in Riverdale, not news. If four people get shot over the weekend in Beverly Hills, news.
You don't even have to try to bias news for it be biased. (Now imagine if you try.)
Underreporting on violence in certain communities vs others is biased. Who determines what is uncommon vs common? Who decides the value of 4 people in Riverdale vs 4 people in Beverly Hills? Just going by “frequency of shootings” is a form of bias.
We can get within a few fractions of a percent of absolute zero but we can't sensibly agree on what "unbiased" even looks like.
You would still have to pose the right questions or risk dying in ignorance.
People's complaints here about bias is really about being manipulated successfully by parts of society working actively in their own, markedly different, interests.
Those won't just go away. Aligning interests doesn't happen by providing facts.
I realise now that my messaging with this project is off.
What would more accurately explain the purpose of this project is to reduce the effects of sensationalism and indoctrination in the news.
In other words, provide a platform for readers to make up their own opinions on news stories by filtering out the excessive positive/negative language.
Even if NewsNotFound articles are only 90% neutral, that's 90% BETTER than reading the same story from a sensationalist source.
Humans provide their sentiment about developments to further their individual interests. Other people want to be informed about those motivations. Taking that away makes news subjectively meaningless for most.
There is a minority of truth-seekers who essentially individually engage in OSINT activity (if only by consuming more than one source).
An AI "doing the thinking for them" isn't the right way though (trust is an issue, remember). You want to teach people, how to be an intelligence analyst and give them the tools to avoid the drudgery.
The goal or call it “aspiration” to serve unbiased news is admirable to me, even while my rational mind knows the aspiration is an unattainable ideal. I will give them a try. Then I can judge their sincerity and performance relative to their aspiration.
Edit: added “judge”
Who defines what is “unbiased”? What are their biases?
You cannot measure something that exists in a fiction that resides independently in each person’s mind.
Objective reality exists, but objective observers do not.
Objective observation is really hard, but amenable to systems and rules that help get closer to this ideal. I'm convinced it's possible to do far better than existing news organisations do just through approaching it with an engineering mindset.
To be even 90% unbiased is 90% better than a sensationalist source.
I'm sure we can get very close to 100% as LLMs progress and if more talented people contribute to the project.
There is no way to judge whether or not you are close to some objective view of “unbiased news”. So it is not a worthwhile pursuit.
As others have noted, what is “newsworthy” is an entirely subjective prioritization of events and information. Are four deaths in Riverside news? Are four deaths in Palm Springs? Are both? Are neither? It depends on your subjective priorities, as a consumer, and no single news organization is ever going to achieve universal, subjective, but somehow unbiased news prioritization.
Labeling anyone person or group’s prioritization of information as “truth” is a dubious endeavor. Dubious to the point of being laughable.
Back to the topic: we seem to be talking about a human being choosing between competing news sources. As that consumer, you get to define "unbiased" for yourself. Wise people have given it a lot of thought, so you might consider consulting what they said.
And you can draw the line at something being measurable to matter to you. That's your choice. I have found the things that matter most are fictions in my brain: love comes to mind.
It is also not necessary to measure to detect a distance. I love my dog, and I love my child. I cannot measure love, yet I detect a significant distance between my love for my dog and my love for my child.
Like love, "bias" has meaning even if it cannot be measured. Like love, one can detect distances between biases, even if they cannot be measured. Science (measurement), math (quantity), and logic (relations of ideas) are just as ephemeral, because they are founded upon unprovable axioms and norms.
I'd love to hear your thoughts after giving the site a try (please use the contact page on the website).
It is by no means perfect right now, but the goal is to get as close to perfect as possible.
Even if NewsNotFound articles are only 90% neutral right now, that's 90% BETTER than reading the same story from a sensationalist source.
If you read sources bias in both directions, you at least know the truth is probably somewhere within that range.
Granted that range can be quite large at times.
What would more accurately explain the purpose of this project is to reduce the effects of sensationalism and indoctrination in the news.
In other words, provide a platform for readers to make up their own opinions on news stories by filtering out the positive/negative language.
Even if NewsNotFound articles are only 90% neutral, that's 90% BETTER than reading the same story from a sensationalist source.
Also, filtering out the positive and negative language might hide the bias of the article, making it seem more authoritative and believable. Many times when I read an article, and I see the journalist making conclusions with bombastic and unbalanced language, it is helpful information toward allowing me to detect bullshit and avoid that source.
Anyone claiming to be unbiased, rather than wearing their biases on their sleeves, is lying to you, or themselves, or they are just ignorant.
Likely some combination of all three.