In response to
"Plus, the algorithms used by Alphabet, Meta, etc have a clear outcome of guiding people down the rabbit-hole."
by
Roger More
|
This is why I think we just need to have regulation specifically about algorithmic recommendations. You can keep S230 in place with targeted
Posted by
oblique (aka kkuphal)
Mar 20 '23, 10:34
|
regulations like that. When the outcomes of algorithmic recommendation are under scrutiny you can damn well bet those companies would be making significant changes to how they recommend. Instead of maximizing engagement, you'd (potentially) see a shift to recommending quality, or helpful, or whatever. Same way they can tune the large language models to be benevolent
|
Responses:
|