Making the Clubhouse hallway more relevant with machine learning
21–27 of 27 posts
Re: Making the Clubhouse hallway more relevant with machine learning
#22Looks like we're at the "personalized recommendations based on extensive user behavioral tracking" phase now. I think every aspiring social app gets here eventually and in my opinion it's ultimately what ruins the whole thing.
> Looks like we're at the "personalized recommendations based on extensive user behavioral tracking" phase now. I think every aspiring social app gets here eventually and in my opinion it's ultimately what ruins the whole thing. Yeah, personalization like that is one of those things that seems superficially appealing but at best just undermines the tool's usefulness with a solipsistic tendency.
Beautifully put. The idea that a social network should be personalised is.. well it is as crazy as expecting any social setting to be personalised, really.
Re: Making the Clubhouse hallway more relevant with machine learning
#23Clubhouse reminds me of 2018 when a lot of PR went into trying to make voice search and Alexa Skills a thing. Of course Twitter Spaces took the wind out of their sails, but what purpose does Clubhouse serve besides being a broadcast megaphone for the already internet-popular 5% of people. Is everyone else supposed to just be the audience, or attempt to chase clout themselves?
Re: Making the Clubhouse hallway more relevant with machine learning
#24I'm happy that the Clubhouse format hasn't caught on.[1] - I hated listening to discussions at specific times and having to schedule my time around it, missing out when something happened in another time zone, etc. - I hated not being able to rewind when I misunderstood something, not being able to pause and think, not being able to return to something afterwards for reference. - I hated the inaccessibility of it. Fo…
This is very much because of your specific social network, and I'm wildly intrigued to know more general notes about it out of curiosity.
Re: Making the Clubhouse hallway more relevant with machine learning
#25Earlier quoted context omitted.
Twitter released their Clubhouse clone, Spaces. Serves essentially the same function, but is native to Twitter so one less app to download. Hosts have incentive to use Spaces because Twitter allows accounts with larger followings to charge admission to Spaces. Clubhouse could be at feature parity now, but for existing Twitter users, there's very little reason to start replicating their networks on Clubhouse.
Clubhouse tanked before Twitter spaces were a thing Both companies missed the mark here
It can be argued that Clubhouse didn't even totally miss it, though I have doubts about their long-term.
Anyway, who do you think did hit the mark?
Re: Making the Clubhouse hallway more relevant with machine learning
#26I'm happy that the Clubhouse format hasn't caught on.[1] - I hated listening to discussions at specific times and having to schedule my time around it, missing out when something happened in another time zone, etc. - I hated not being able to rewind when I misunderstood something, not being able to pause and think, not being able to return to something afterwards for reference. - I hated the inaccessibility of it. Fo…
> I know about Twitter Spaces, but I haven't come across any yet. When Clubhouse came, suddenly everything around me was happening on Clubhouse. This is very much because of your specific social network, and I'm wildly intrigued to know more general notes about it out of curiosity.
For a few weeks ClubHouse became very popular around here and major serious media, NGOs, etc. were all creating content there (instead of putting it e.g. on their YouTube). That's why it got me worried.
During that time I sometimes tuned into some US rooms as well (e.g. Jason Fried hosted a regular and interesting room), but I always thought that it would work better as a podcast.
Re: Making the Clubhouse hallway more relevant with machine learning
#27Hello whomever from clubhouse is promoting this blog post on the internet. Looking forward, please consider adding more technical depth to the content. This is meant to be a technical blog post and the only technical description of the system/model pipeline involves telling us its a GBDT. That's unfortunately probably the least interesting thing you could pick out to say about such a recommender. The content briefly…