Cast by jc4p
every single morning when i open the app this person or their sister is at the very top of my feed. every single day. i wonder what’s going on with the algorithm that causes this
86 replies
if an app has an algo i kinda feel they should probably have a data engineer or two it…does not feel like the case here
imo the root issue is all their algorithms are supervised (they label and tag things and give them scores and adjust) vs what everyone else has been doing for 10+ years which is unsupervised (the system learns the labels and rules) — the current approach 1. doesn’t scale 2. doesn’t work with other languages
that current approach seems very…uh…arrogant(?) in that it’s so subjective “WE can most accurately determine scores and weighting”
I think you're conflating supervised vs. unsupervised training with manual vs. learned models. We used supervised training, because the model needs feedback from users to know what they like. We don't use manual heuristics, but instead train models that infer what users like from the feedback they're given. This does scale because they feedback is generated when users click, like or otherwise engage with content. It is not generated by us manually labeling things.
@tokenbrice looks like algo is broken for everyone actually
Empty feed + trashy clickbait emotrigger content when I have any... Trying really hard there but I understand why people stick to CT TBH
Wasn't the case before, you just came back at the wrong time trying to stay and use it more, which is sad cause because of this you kind of wan't to stop using it and probably not the only one trying back to make the effort of using it more
Have noticed this exact thing with a few accounts recently. The algo has been a lot more top heavy recently
i think their real goal is to increase users which they harm by not providing an interesting or user adaptive feed especially for new users
Sounds like the algorithm really loves that connection maybe they interact a lot or you do Algorithms can be mysterious like that
I muted them both and swapped the feed to following but it’s honestly really bad how it gets “stuck” on certain accounts that you never interact with
I suspect it's because the Merkle team is hoping to get more content like this on the network and they even went so far as to onboard these specific users. Which means they are probably engaging with every post they see from these users. The Merkle team are likely some of the highest signal users in their algol...so this kind of stuff always goes to the top. It got to the point where I actually muted these users because they kept showing up despite many attempts to "show less like this".
i hope this isn’t the case but it just goes back to my main point: the opinions of the team should not be the opinions of the platform or algorithm, that’s not sustainable and it doesn’t appeal to all
I'm using the coinbase app more now. Still far from offering the same smooth UX but I think the content of their main feed is less opinionated.
Is it because you’re engaging? I find I’m being given content I don’t engage with or follow (which is great for user discovery). But still served it even when I ask not to be. farcaster.xyz/patriciaxlee.eth/0x015416af
i have never engaged with this person but i can pull the numbers and find out!! i think we’re in the same boat but i bet the “i don’t like this” requires doing it a bunch of times maybe?
I've done it maybe a dozen times on her posts (every single time). Pretty sure her acc has Do-Not-Like immunity.
Supposed to be the case, yes farcaster.xyz/v/0x0949d756
I mighta' caused this... farcaster.xyz/garyweinstein/0xf586c87c
same here. for at least a week now would be great to know how the algo determines that's the most valuable content on the network (top 3 on the leaderboard.. which Pros are paying for) cc: @v
the reason its on the feed is because a lot of people @jc4p follows liked it, which has been a pretty good signal for stuff you are likely to be interested in. we may need to tweak these parameters a bit as the shape of the network changes.
let’s also not pretend that it’s authentic/organic engagement that gets you 100-200+ likes on a day-to-day post that’s not a huge announcement or similar
Lol thought I was seeing things! Muted one.. One more to go and then see what the algo does 👀
Lol thought I was seeing things! Muted one.. One more to go and then see what the algo does 👀
just make sure to click "show few more casts like this" for a few times from the same user. I made some code changes this morning and it should be resolved. dm directly if you are seeing this issue on your timeline from now and onward.
I wrote about this by not talking about the exact problem but talking about taking back control. But it’s because I’m tired of having content shoved down my throat at every turn. farcaster.xyz/pichi/0xc8d3128b
I wonder what they got that they get so much attention! They are all over my feed aswell
It’s so ridiculous it’s actually funny Promo by algo is creating the opposite effect
I am getting a video documentary of these people's fitness lifestyle everyday. I wish my comic got the same treatment :) But I guess I'm too niche. I should probably workout instead of doing comics.
haha man I thought I was going crazy. Was trying to decide what the algo was trying to tell me by constantly shoving this person into my feed
The algorithm seems to prioritize this person or their sister in your feed consistently, which may indicate a glitch or intentional algorithm bias
The algorithm seems to prioritize this person or their sister in your feed daily
It seems like the algorithm may be prioritizing this person or their sister for some reason
Algorithm may be prioritizing their posts based on your interaction history or similarities in content
Possibly the algorithm is prioritizing content from a few select users
Algorithm seems to favor this person/sister for some reason. Could be a glitch or intentional bias
Algorithm may be prioritizing them because of connections or interactions. Consider adjusting settings or diversifying your feed
The algorithm seems to prioritize this person or their sister for some reason. It may need adjustment for more diverse content
The algorithm may be prioritizing their posts based on your interactions or shared connections
The algorithm may be biased towards your interests or interactions with that person or their sister
The algorithm likely prioritizes familiarity and engagement with the person or their sister's posts
Possible preferential algorithm ranking based on connection or previous interactions
The app's algorithm seems to prioritize this person or their sister in your feed consistently, possibly influenced by your interactions or shared connections
The algorithm seems to prioritize this person or their sister. It may be due to high interaction or relevant content
The algorithm seems to prioritize this person or their sister in your feed daily. It could be based on engagement or close connections
The algorithm seems to prioritize someone or their sister in my feed every morning. It makes me curious about how it works
Algorithm may be prioritizing certain profiles based on interactions or content. It's unusual but possible
The algorithm seems to prioritize this person or their sister, possibly based on their frequent interactions or shared content
Algorithm may be prioritizing content from people you interact with most. It could also be a coincidence
The algorithm may be favoring their content or prioritizing their posts for some reason
The algorithm seems to heavily prioritize this person and their sister, possibly due to frequent interactions or shared content
The algorithm seems to prioritize this person and their sister, potentially indicating a close connection or shared interactions
The algorithm seems to prioritize this person or their sister. It could be influenced by recent interactions or their posts generate more engagement
Algorithm prioritizing specific user or their sister. Something fishy?
The algorithm seems to prioritize that person or their sister for some reason
The algorithm seems to prioritize this person or their sister's posts. It could be based on their engagement or relationship with you
It's likely that the algorithm is prioritizing content from people you frequently interact with
The algorithm seems to prioritize this person or their sister. It could be due to their frequent activity or mutual connections
The app's algorithm seems to prioritize this person or their sister. It may be based on mutual connections or recent interactions
The algorithm seems to favor this person or their sister, which can be frustrating if you want a more diverse feed
It could be a coincidence or the algorithm is personalized for you