But perhaps the more important filtering is on quantity as opposed to neutrality? Perhaps filtering out a large amount of news, even with some bias, is the lesser evil, as compared to news outlets that depend on stirring the emotions of their readers every single day?
Wikinews used to be okay in this regard, but the German version I used has died down a bit, and the English one is even more centered on the Anglosphere than HN.
Sometimes there aren't multiple sides, especially when it comes to science reporting. You have fact-based reporting, and then you have conspiracy theories.
How would you handle news where there is sufficient evidence to show one set of reporting is accurate and relatively unbiased, but another report is all made up and designed to inflame its audience?
In this example it sounds like there are multiple sides, just that one is baseless. Even these though I have found come from somewhere, maybe a misunderstanding or a conflation of unrelated topics. While not unlikely intentional on the author's side, the readers are not so sinister I think. Being able to read this while grounded with the other more factual side helps when discussing with those I may otherwise generally disagree with. It has felt somewhat like language learning, understanding it helps with communication with people with a very different background.
Admittedly it takes more time to do this, and I can see not being able to invest that in a general sense. I personally think it's worth it.
That's rarely the case with science reporting. The subjects that are sufficiently rigorous to allow no reasonable debate (the physical sciences) are rarely political enough to inspire unreasonable debate.
On the other hand, the subjects that are politically contentious are not rigorous and leave plenty of room for reasonable debate.
If anything, science reporting tends to err the other way, uncritically reporting sensational results that contradict one other, have not been confirmed, or fail to replicate.
I rarely see a popular science article that doesn't report the results of a single experiment as if they were instantly established fact.
I agree with what you say about sensationalist science reporting. It's very common to take one study and then have dummies who aren't scientists report on it as if we've just found out how to live forever. Science is very tricky because it's complicated and the barrier to entry is high - you can't just extrapolate things out like that.
However, there's also the other side of things, which is mostly established science. Which, you're right, don't typically spark political debate... but they do sometimes. Vaccines, climate change, cholesterol, seed oils. The RFK Jr faction of anti-science is rife these days.
I'm not familiar with all the details of RFK Jr's position, the public debate about it, nor how many of the things reported in the news are accurate (truth being the first casualty of politics).
But I just looked for those topics in the official Make America Healthy Again report [1].
The positions in that report on those topics were not so unreasonable. It says seed oils are a concern because they are ultra-processed fats, only mentions cholesterol in the context of PFAS, and says "vaccines benefit children by protecting them from infectious diseases" but we may not need to give children nearly 30 doses of them.[2]
I do think his general position that processed food is unhealthy is not only reasonable, it generally matches conventional modern medical thinking, even if he is wrong about a few details.
And of course people looking for political ammunition only look for details they can use against him.
I think you may be on to something here. When it comes to news, none is bad but so is too much. This would be true even if all the consumed news was politically neutral and completely objectively factual and accurate. But of course all news is biased, and much very deliberately so to the point of obscuring the truth of it. LLMs are not going to make this situation better.
I want to be aware of what's happening, but not to drown in it. How to achieve that is not only a good question but the right question.
But perhaps the more important filtering is on quantity as opposed to neutrality? Perhaps filtering out a large amount of news, even with some bias, is the lesser evil, as compared to news outlets that depend on stirring the emotions of their readers every single day?
Wikinews used to be okay in this regard, but the German version I used has died down a bit, and the English one is even more centered on the Anglosphere than HN.