Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

YouTube’s ‘dislike’ barely works according to new study on recommendations

engadget.com

Post reply on HN
Print view View original

431–437 of 437 posts

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#431

Post by brigandish »

autoexec wrote: post #413
brigandish wrote: post #409

Earlier quoted context omitted.

I use DDG now but I wish the search tools weren't just last week , last month , last year but greater than a week/month/year so I could discard more recent results. It's quite the puzzle to find ways to remove things, as using the -search-term-I-don't-want or +term-I-do-want is increasingly less reliable, in my experience.

Yeah, google pretty much ignores search modifiers these days and DDG (which I also use) routinely fails to understand what -term means. I still dream of the day we have a search engine that supports regular expressions.

There was one, back before the great clearing of interesting search engines that Google brought about. I forget its name but I used it for code searches.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#432

Post by dredmorbius »

dredmorbius wrote: post #371
daveoc64 wrote: post #131

Earlier quoted context omitted.

The problem with the dislike button seems to be that nobody agrees on what it should be used for. In previous HN discussions, it's become apparent that some people use it to say "I am not interested in this video". That is of zero use to other users. The video isn't necessarily bad, it's just on a topic you don't care about. If you are instead using it to mean "this video is factually wrong" or "this video is obnoxio…

In the spirit of POSIWOD, I suggest POBIHOU: the purpose of a button is how it's used. Designers may have an intention. People will apply their own needs and usage to features, in the spirit of desire paths. https://en.wikipedia.org/wiki/The_purpose_of_a_system_is_wha... > https://en.wikipedia.org/wiki/Desire_path > If people are using "dislike" to indicate that they're not interested in content, particularly in a co…

That should be POSIWID and POBIHIU, I suppose...

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#433

Post by soundnote »

ra120271 wrote: post #283
adolph wrote: post #232

Earlier quoted context omitted.

There is something to be said for a search engine whose interests don't align with, or even better are adversarial to, the prevailing spam sources.

I used to love that about Google News in the early days. A search would give you results from all over the world, with different outlooks and perspectives. I found it enlightening to read against my own selection bias. Then they removed that value point and aligned what they presented with my existing bias - at which point I stopped using it.

I like Brave's Goggles idea on that front, and for other stuff as well, like curating knowledge domain-specific site boosting lists.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#434

Post by soundnote »

esotericsean wrote: post #242
dijonman2 wrote: post #221

Earlier quoted context omitted.

Better than bing image search, my current go to?

Someone uses Bing?

Just about anyone not using Google (Startpage) or Brave uses Bing, since it's the source of just about all indie engines' results. And even Brave's image results come from Bing for now.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#435

Post by Sophistifunk »

kilroy123 wrote: post #91
Sophistifunk wrote: post #75

Why is everybody pretending the recommendation system is designed to show you what you want to see, rather than to maximise ad revenue?

I assume this because I pay and don't see ads.

So do I. As such, what we watch doesn't make much difference to their bottom line, so they're certainly not going to write a second, just-as-expensive-to-maintain recommender engine just for the three of us.

Re: YouTube’s ‘dislike’ barely works according to new study on recommendations

#436

Post by ehsankia »

behaveEc0n00 wrote: post #40

The zeitgeist of our culture; of course you didn’t MEAN to be negative, here try again. No, no; I most certainly did mean to suggest you should stop showing me this persons channel. Not because I hate the person but because I don’t want a shit “AI” that’s feeding me what Google makes the best margins on.

What I find interesting is the duality of YouTube criticism. On the one hand, YouTube should respect people's like and dislikes to a letter, but also at the same time YouTube pushes people into dark echo chamber rabbit holes by only giving them what they want.

On the the one hand they need to stop accidentally taking down innocent videos, but also need to take down bad videos faster.

People don't realize that all these issues are two sides of the same coin. They complain on each side separately and don't understand why optimizing for one messes up the other.

Post reply on HN