Good evening. Can we fix this with our own feed algorithms "store", transparency, and user choice on Nostr? I feel algorithms aren't inherently bad, that they can do a lot of good for content discovery and user discovery, it's just that what we have today on other platforms isn't very transparent, closed, and lacks any user choice. We can do better.
Discussion
noogle.lol/discover
All NIP90 DVMs, usable by any client that implements the NIP. Content algos, both free and subscription based. Would love to see more people building and sharing their own.
Content discovery doesn't have to rely on one evil algo decorating our feed, giving power to whoever controls the algorithm. Content discovery should be a matter of choice for users, ideally in an open source way so that everyone can see why stuff gets shown to them, and if we don't like how it behaves we pick another one, or even write our own.
Yes. If we own our own data, then we can own our own feed, too.
should it be separated into NIP90 content creation and NIP XX content discovery via algo?
I don't really understand that question π
put these in a different nip kind 5300.. https://www.data-vending-machines.org/ranges/53xx/
May also be interesting to have a toggle that allows you to flip your default algorithm to something close to the opposite of what you normally see. Like a βDark modeβ for content that flips to the inverse of your normal interactions or topics. Maybe need some AI to generate that. Just an idea.
You could still have content filters apply, but look for posts outside your social network normal interactions.
maybe an inverse Web of Trust? Web of Distrust? Further away, the higher the score
Yes! Though I still will not want to see violent or abusive content show up in my timeline if there was a way that I had chosen on the client to filter out those tags (especially if little ones are around).
Due to current lack of client side user choice filtering I hesitate to invite a whole group of others to this blossoming new technology. I know many moms who will not abide by such lack of feature. Maybe this is part of why womenβs donβt feel comfortable here?
We want a permissions web, but the user should be able to withdraw permission from being exposed to known content types at least by tag (IMO).
yeah, i agree... it's an extremely raw experience at this point... the potential for newbies to rage quit after some fuckstain posts dickpics and said lady was using a client that didn't filter it out because lack of WoT filtering...
WoT solves so many problems with this with onboarding, you just start from one, your referrant, and after that WoT does much of the heavy lifting
usually a friend is not so distant in their tastes to a person, that it's going to be a rage quit situation
and actually i think that WoT is very mom friendly
i think over the next year there will be a lot of advancement in the area of decentralised, subjective filtering techniques, so keep poking at it and figuring out how to do it
nostr:npub1sg6plzptd64u62a878hep2kev88swjh3tw00gjsfl8f237lmu63q0uf63m understands better than anyone what we are building here
it is the town square, the forum, the marketplace of ideas
you don't put the groceries stalls next to the whores, what can i say? bad architecture
yes, i saw it in Canberra back in the day: you put the whorehouses next to the hardware stores and construction businesses, just sayin