18
votes
Google, Microsoft, and Perplexity promote debunked scientific racism in AI search results
Link information
This data is scraped automatically and may be incorrect.
- Authors
- David Gilbert, Matt Burgess, Brenda Stolyar, Virginia Heffernan, Parker Hall, Reece Rogers, Paresh Dave, Will Knight, Kate Knibbs
- Published
- Oct 24 2024
- Word count
- 1765 words
I tried getting a summary of the top links in Kagi out of curiosity and got this:
Unfortunately it seems like the problem is that the top search results are racist websites that look legitimate. It reminds of a story a professor told about a student that wrote a paper about how awful some Black civic leader was citing the white supremacist site Stormfront (Wikipedia). The student wasn't racist. He'd just clicked some of the top search results and trusted what he found on the internet. LLM summarizers do a lot of the same. Racist garbage in, racist garbage out...
This part at the end of the article is pretty telling, I think. Yes, it's bad that these AI systems are parroting garbage data from white supremacists. But with the sheer number of human academics who have been willing to parrot the same garbage data from white supremacists, it's not exactly shocking.
We've had those people posted here and their opinions highlighted or glossed over. I refuse to wash those folks' opinions anymore.
Mirror: https://archive.is/z10NQ