In theory, but I think the problem you'd find in practice is that reviewing 100s of nearly-identical AI-written articles gets very boring very quickly, and a lot of errors would slip through.
Ok, so 2 is not enough. What if you put 4 humans? Or 8? Or maybe even 20 (instead of 100, per OP)?
These experiments fail because they greedily try to outsource all the work to AI. Another recent example: the lawyer who submitted ChapGPT hallucinations directly to a court case.
You don't need to eliminate humans, and certainly not at first. You just need to be much more efficient than status quo in order for AI to be deployed at scale.
Why? Lots of people do a lot of reading for a living. Reading a new article is way more interesting than thousands of jobs I can think of. Data analysts and accountants literally pour though millions of featureless numbers over their career, why would this be any different?
It's like trying to find typos in your own writing. It's very difficult to stay focused when reading a long series of nearly identical documents. The people you're talking about are reading a lot of novel documents.
> Why would these be identical? New stories everyday, new games played etc.
If the only data you have about a game is the sport, the team names, and the score, there's only so much an AI (or a human!) can do to write an interesting article about it. Once you've read a few dozen of the generated articles, they'll all start sounding the same -- because, aside from the details, they are all the same.
If you want quality articles, you need some more depth in the source data. And, for little-league sports games, that data may just not exist.
> One article even failed to populate properly, with the text instead featuring a bracketed glimpse at how its opening sentence was supposed to read.
> "The Worthington Christian [[WINNING_TEAM_MASCOT]] defeated the Westerville North [[LOSING_TEAM_MASCOT]] 2-1 in an Ohio boys soccer game on Saturday," reads the butchered intro.
These experiments fail because they greedily try to outsource all the work to AI. Another recent example: the lawyer who submitted ChapGPT hallucinations directly to a court case.
You don't need to eliminate humans, and certainly not at first. You just need to be much more efficient than status quo in order for AI to be deployed at scale.