4 ms·
The article offers no new or inside information, just more armchair quarterbacking. I'm surprised that it is getting traction on HN. I think it says more about
by xibalba 5y ago
The article offers no new or inside information, just more armchair quarterbacking. I'm surprised that it is getting traction on HN. I think it says more about the zeitgeist than it does about the (lack of) insightful-ness of the content.
- gwern 5y agoYeah, it just repeats the Narrative in a giant post hoc. "Zillow uses ML models in some way; Zillow failed; QED, ML models are dangerous." Except the reporting by Bloomberg and insiders is that Zillow failed because they overrode the models predicting lower prices and bought like drunken sailors, and it's just a story of yet another marketmaker being run over by the market. So sad, too bad, largely irrelevant to the tech world, yet, it looks like it's entering the mythology of ML up there with Cambridge Analytica or the tank story - unkillable by mere facts or tardy reporting.
- adjkant 5y agoAs someone who upvoted but didn't care much for the content, I think it's worth mentioning that sometimes/often I upvote for the currently occurring conversation to get more eyes, not for the link. I haven't seen too many specifics on the Zillow collapse and I've learned a nice deal through many of the comments here, most not having much at all to do with the article.