Posted by ninjagoo 6 hours ago
They will announce official paid AI access plans soon. Bookmark my works.
Sell a "truck full of DAT tapes" type service to AI scrapers with snapshots of the IA. Sort of like the cloud providers have with "Data Boxes".
It will fund IA, be cheaper than building and maintaining so many scrapers, and may relieve the pressure on these news sites.
But - as another poster pointed out - Wikipedia offers this, and still gets hammered by scrapers. Why buy when free, I guess?
> just like how the agricultural sector is hell-bent on scapegoating AI (and lawns, and golf courses, and long showers, and free water at restaurants) for excess water consumption when even the worst-offending datacenters consume infinitesimally-tiny fractions of the water farms in their areas consume.
When I learned about how much water agriculture and industry uses in the state of California where I live, I basically entirely stopped caring about household water conservation in my daily life (I might not go this far if I had a yard or garden that I watered, but I don't where I currently live). If water is so scarce in an urban area that an individual human taking a long shower or running the dishwasher a lot is at all meaningful, then either the municipal water supply has been badly mismanaged, or that area is too dry to support human settlement; and in either case it would be wise to live somewhere else.
But then it was not really open content anyway.
> When asked about The Guardian’s decision, Internet Archive founder Brewster Kahle said that “if publishers limit libraries, like the Internet Archive, then the public will have less access to the historical record.”
Well - we need something like wikipedia for news content. Perhaps not 100% wikipedia; instead, wikipedia to store the hard facts, with tons of verification; and a news editorial that focuses on free content but in a newspaper-style, e. g. with professional (or good) writers. I don't know how the model could work, but IF we could come up with this then newspapers who have gatewalls to information would become less relevant automatically. That way we win long-term, as the paid gatewalls aren't really part of the open web anyway.
Journalism as an institution is under attack because the traditional source of funding - reader subscriptions to papers - no longer works.
To replicate the Wikipedia model would need to replicate the structure of Journalism for it to be reliable. Where would the funding for that come from? It's a tough situation.
The Wikipedia folks had their own Wikinews project which is essentially on hold today because maintenance in a wiki format is just too hard for that kind of uber-ephemeral content. Instead, major news with true long-term relevance just get Wikipedia articles, and the ephemera are ignored.
Interesting idea. It could be something that archives first and releases at a later date, when the news aren't as much new
Practically no quality journalism is.
> we need something like wikipedia for news
Wikipedia editors aren’t flying into war zones.
Which is a valuable perspective. But it's not a subsitute for a seasoned war journalist who can draw on global experience. (And relating that perspective to a particular home market.)
> I'm sure some of them would fly in to collect data if you paid them for it
Sure. That isn't "a news editorial that focuses on free content but in a newspaper-style, e. g. with professional (or good) writers."
One part of the population imagines journalists as writers. They're fine on free, ad-supported content. The other part understands that investigation is not only resource intensive, but also requires rare talent and courage. That part generally pays for its news.
Between the two, a Wikipedia-style journalistic resource is not entertaining enough for the former and not informative enough for the latter. (Importantly, compiling an encyclopedia is principally the work of research and writing. You can be a fine Wikipedia–or scientific journal or newspaper–editor without leaving your room.)
- crowdsourced data, eg, photos of airplane crashes
- people who live in an area start vlogs
- independent correspondents travel there to interview, eg Ukraine or Israel
We see that our best war reporting comes from analyst groups who ingest that data from the “firehose” of social media. Sometimes at a few levels, eg, in Ukraine the best coverage is people who compare the work of multiple groups mapping social media reports of combat. You have on top of that punditry about what various movements mean for the war.
So we don’t have “journalist”:
- we have raw data (eg, photos)
- we have first hand accounts, self-reported
- we have interviewers (of a few kinds)
- we have analysts who compile the above into meaningful intelligence
- we have anchors and pundits who report on the above to tell us narratives
The fundamental change is that what used to be several roles within a new agency are now independent contractors online. But that was always the case in secret — eg, many interviewers were contracted talent. We’re just seeing the pieces explicitly and without centralized editorial control.
So I tend not to catastrophize as much, because this to me is what the internet always does:
- route information flows around censorship
- disintermediate consumers from producers when the middle layer provides a net negative
As always in business, evolve or die. And traditional media has the same problem you outline:
- not entertaining enough for the celebrity gossip crowd
- too slow and compromised by institutional biases for the analyst crowd, eg, compare WillyOAM coverage of Ukraine to NYT coverage
Isn't that what state funded news outlets are?
I wonder if bots/ai will need to build their own specialized internet for faster sharing of data, with human centered interfaces to human spaces.