It’s conspiracy theory to suggest without evidence that “anti-narrative” information is being “removed”. Yes, big tech is certainly shaping results, but they’re doing that in both directions. They aren’t simply suppressing data they (or the hypothetical overlords) disagree with. They are feeding the info that people want. (Which is still bad)
The existence of that article in the New York Times would certainly seem to indicate a lack of suppression.
I do appreciate the link, though. That’s interesting and I hadn’t heard that was a concern. I wonder if there is science evidence that or if it’s hypothetical. The article doesn’t indicate one way or the other unless I missed it.
I will note it took me some time to find the article (I knew it existed because I'd read it when it came out). Google very nearly refused to do anything but send me links about how important it is to get boosters.
I don't know a ton about the science behind it other than some teams in Israel seem to have pretty strong feelings about it (more pointed than what's in the NYT article), but in my quick search I didn't see them in the search results.
With regards to search shaping, it's pretty easy to see: just Google image search "black inventors" then "white inventors". Also try "black family" then "white family". I'm not claiming any kind of oppression here, just noting that one of the sets of results looks like a Benetton-style diversity ad, while the other just has black people. I find it difficult to believe this was an organic result that wasn't explicitly influenced behind the scenes.
> With regards to search shaping, it's pretty easy to see: just Google image search "black inventors" then "white inventors". Also try "black family" then "white family". I'm not claiming any kind of oppression here, just noting that one of the sets of results looks like a Benetton-style diversity ad, while the other just has black people. I find it difficult to believe this was an organic result that wasn't explicitly influenced behind the scenes.
You could be right. I certainly see that “white family” involves a fair number of pictures of non-white folks. Individually the results all make sense (one from an article called “my white family”, stock photo “interracial black white family”, another about from an article about adopting a white child), but it’s odd that “black family” doesn’t have the same.
At the same time, it’s entirely plausible that this is just surfacing biases in the input. Maybe articles about families that aren’t just white have higher page rank? Maybe it’s something else. It’s interesting, but I’m doubtful it’s intentional (but it could be). If Google wanted to push a bias, it would make a lot more sense to push it on the unqualified “family” or “inventors”.
“White inventors” showing some black inventors makes a lot of sense given popular articles like “The iconic American inventor is still a white male” that specifically discuss non-white inventors.
The existence of that article in the New York Times would certainly seem to indicate a lack of suppression.
I do appreciate the link, though. That’s interesting and I hadn’t heard that was a concern. I wonder if there is science evidence that or if it’s hypothetical. The article doesn’t indicate one way or the other unless I missed it.