https://en.m.wikipedia.org/wiki/Russian_invasion_of_Ukraine
Problem is we need the same for local news. A place where all the research gets archived, and events are tracked. Seems like nowadays the only people with a complete-ish picture of what's going on at a local level are the people directly involved themselves.
Anywho, things aren't looking good. One of the media giants in my country just cut 4,800 jobs across the country:
https://www.cbc.ca/news/canada/british-columbia/david-eby-br...
----
January 2023, I finally had an LLM cite one of my own written contributions when I began asking about a nuanced technical database which I maintain the article on wikipedia (which was really cool/eerie). This was over a year ago, and every day I continue to be floored by genAI capabilities in discussing published dataset complexities.
I know several lowkey published authors (one whose autobiography I copyedited) that have been terrified by what LLMs know about their own lives, based on commonsense associations/deductions.
The prompt had not specifically mentioned either HN or my username.
This makes me think that the depth of training material on at least certain topics is thinner than might be hoped.
What you end up getting is the approved narrative of "how things are" by the same class of people who all hobnob with each other.
Glad to see it's end. Look forward to what's next.
It also seems like journalism is suffering a similar problem as software engineering, in that people think they can take shortcuts. "I'll just walk around and ask people questions" is not the same thing as journalism, just as "I'll write a To Do app" is not the same thing as doing software engineering. These are disciplines with exacting standards, quality levels, and rigor, and require years of diligent study and practice.
And in the case of well-practiced journalism, it really is "the fourth estate", and we should be fearful of what our world becomes should that fourth estate ever crumble.
This was Vox's initial schtick, but they kind of abandoned it after a while- it didn't bring readers in, and they wanted to be ad-supported.
P.S. if anyone wants to give me a million dollars to do this, I am both an experienced news editor and a senior-level software engineer, email in profile
to get people to see how this could make journalism more useful and powerful
Did you ever consider news media is structured the way it is by design? It's inherently a distraction and engagement tool crafted to influence public perception to the whims of those behind the companies.What you propose is nearly the opposite of that and would likely incriminate some powerful actors.
Hypertext and databases were made for this, but the news media can’t see it. I tried, for years, in school, at conferences and in various journalism jobs, to get people to see how this could make journalism more useful and powerful, but here in 2024, we’re still doing “articles, except on the internet”. And I can’t afford to keep working in the news media long enough to be in a position to try it out. Nobody can, unless their actual livelihood is coming from somewhere else.
So here we are, in total context collapse, and a dying news media whose slow disappearance, especially at the local level, is actively hurting people, and our democracy.
It’s bad, and its going to get a lot worse as AI makes it cheaper and cheaper to lie convincingly. But what can anyone do?