5 ms·
> It's inordinately effective where algorithms promote material based on short-term appeal rather than long-term benefit. Could propose a way to measure long-t
by freemint 5y ago
> It's inordinately effective where algorithms promote material based on short-term appeal rather than long-term benefit.
Could propose a way to measure long-term benefit?
If you have an proposal for a metric, can it be used to make decisions which video to show now, without needing to access to information from the future?
- dredmorbius 5y agoWhere is the clue and creator talent going? Is current usage an indication of attraction (appeal) or of frustration (lack of a viable alternative)? That said, the problematique of distinguishing short-term and long-term benefits (or harms) is a difficult one. It's inherently one that markets fail at fairly spectacularly, though so do numerous other systems. I'd ... expect ... that there's some prior literature on this in the media space. It's also in large part the domain of aesthetics in philosophy. There are a few institutions which have managed to retain a high quality standard over time. At the scale of the Internet, I'd nominate HN, and largely credit its moderation for that. Having principles, values, and an enforcement mechanism for those helps. At a larger and longer-term scale, universities are among the very longest-lived and highly-durable human and social institutions. I'm not sure of all the reasons for this, and it's possible that in some cases this is a historical accident. It's also true that there have historicially been tremendous barriers to entry. Oxford and Cambridge in England / the UK had a chartered monopoly on all higher education until the 19th century. Even after the removal of that protection, they've remained highly durable as elite institutions.