The more I see of social media, the more I'm convinced that it's not really a fulfilling way to share and socialize on the internet. I've come to believe that anything that can platform every voice on the planet forces noise to drown out signal.
Since you've actually started a social media site, I wonder if you have any insight on that? You say you believe in this project; what is it that you think is valuable about letting everyone talk to everyone, as opposed to exclusive and curated spaces for people that semi-self-select based on community interests and values, like forums? Separately, why do you think that removing organizations/brands and psuedo users is going to be the crux of the solution?
Honestly looking to have my mind changed. But mostly just curious.
I like the idea at a social level/product level, but at this point I don't think I'd use any social network that wasn't open at a protocol level in some way. ATProto/etc would be a great tech backbone for this, and would then let users self host their content, it would let orgs run their own views, it would let archivers archive content for posterity, it would open the door to possible federation in ways that suited the network in the future. Re-inventing all of that from scratch would be a terrible idea, and would lose focus.
A “social contract” that one party changes at will and then just announces the changes is not a contract (or social) in any meaningful sense. It’s just terms and conditions with a different name.
If you want people to see it as a different thing, it needs to be a different thing.
Do you have any plans to support Ultra HDR images, on the Android app and/or other platforms? I did some browsing of the site and didn't see any examples (but that doesn't mean it's unsupported).
I actually had to look this up. Currently we don't but apparently we could support it if you upload via the web. Our mobile apps use Expo and the native image picker re-encodes the image.
If you're interested still despite limitations (Ultra HDR Images also won't be seen by many people) I can look into this more.
Thanks for taking a look! I haven't used Expo, but looking at the docs I wonder if using 'allowEditing: false' and 'exif: true' in the ImagePickerOptions would bypass reencoding? The gainmap is stored in the EXIF data. Instagram is similar in that the gainmap is stripped if you edit the image at all, cropping included, but if no crops or filters are done then the image retains the gainmap.
Either way if the clients are able to render the HDR images, I think uploading on web is still a usable implementation.
I just did some more research and we probably won't do the work until there's more demand. We can actually easily support it on desktop (we upload the raw bytes but always default to serving a compressed version, which I can override for HDR) but for mobile there's no drop-in solution for Expo apps and thus would have to build 4 new things (iOS and Android uploader and display components). That said, I'd like to do this because I want PIECES to have a rich visual experience and was thinking of recruiting more photographers, but probably not now. Thanks for your interest though.
Doing it by hand now. Besides the site itself, I have spent most of my time building out the backend automation. I imagine I can manage the first 10k users myself and opening it up slow enough that a community itself develops and will be self-policing.
I got back on instagram recently and the first thing I see when I login now is content from repost accounts and it's like, do I really want to see this? Is this what happens when we optimize for engagement blindly? A hypothesis I have is that if we make rules in the beginning this will be easier to enforce at scale. See an account that's just reposting stuff -> report it -> review
I hope this isn't coming off as snarky or anything, I am honestly interested in the project...
So you're thinking something like Twitter(X)'s community notes, but instead of flagging misinformation, it will be used to 'community flag' an account as AI? Maybe it shadow bans, maybe it brings it to moderation queue, whatever...
I've seen a few of these 'NO LLM ALLOWED' sites pop up now, but any countermeasure to LLM use is easily subverted. I think the only real way to keep the LLMs out is, as you seem to agree, some sort of user-reported system (But that seems open to fraud....difficult problem to solve!)
For sure. AI is only getting better so it's not a promise I can make but it's worth stating from the beginning. But the thing about Meta/Twitter/etc.,. is that they're financially incentivized to promote these accounts (more eyes = more ad revenue) so that makes it even worse. There will never be an explicit financial incentive to run bot accounts/repost farms on PIECES (I'm sure people will try though).
I recognize that we'll have to re-think things once we hit millions of users (I am an optimist). But for the first ten thousand or hundred thousand, I'm confident we can keep it a people-centric place and then strategize for future growth once our strategies start to be less effective.
I'm going to sound a little arrogant here but part of it is because most people don't actually believe in what they're doing. I see posts of founders who write, "hey, we started a social network, but we've since pivoted to <insert AI tool here>".
I believe in PIECES to my core. Everything I'm doing now with my full-time job so I don't have to take outside investment and saving is to get not only PIECES but my dating platform Lettr Dating off the ground. PIECES will be around in 5 years, I guarantee it, because it's the social network that I always dreamed of having - even going back in the day to 2003 when I would just post on my xanga.
This doesn't seem to have been picked up by the HN algo, which is fine. But I'm going to keep at it.
If you're in NYC, I'll be promoting PIECES and Lettr in the parks at night time, and I'll likely be hiring an intern at NYU to help me spread it around campus. If you see this and are at NYU, I'd love to chat - https://piecesof.me/p/6qrjydihiq3tikg3fb85fqcelqnhph
The software is really well-made.
I wonder why you change the links from the internal ones to this format https://piecesof.me/p/5b2w8x6sn2mwkdhppfusvqsf8z5h1t ?
The writing is quite nice. I don't understand the "Docs". And I'd suggest to make it very easy to write my first post. The big editor is a bit scary.
also here: https://piecesof.me/videos - where do I get prompted to upload my own video?
Anyway, good luck! Let's see how long I can stick around.
The more I see of social media, the more I'm convinced that it's not really a fulfilling way to share and socialize on the internet. I've come to believe that anything that can platform every voice on the planet forces noise to drown out signal.
Since you've actually started a social media site, I wonder if you have any insight on that? You say you believe in this project; what is it that you think is valuable about letting everyone talk to everyone, as opposed to exclusive and curated spaces for people that semi-self-select based on community interests and values, like forums? Separately, why do you think that removing organizations/brands and psuedo users is going to be the crux of the solution?
Honestly looking to have my mind changed. But mostly just curious.
I like the idea at a social level/product level, but at this point I don't think I'd use any social network that wasn't open at a protocol level in some way. ATProto/etc would be a great tech backbone for this, and would then let users self host their content, it would let orgs run their own views, it would let archivers archive content for posterity, it would open the door to possible federation in ways that suited the network in the future. Re-inventing all of that from scratch would be a terrible idea, and would lose focus.
A “social contract” that one party changes at will and then just announces the changes is not a contract (or social) in any meaningful sense. It’s just terms and conditions with a different name.
If you want people to see it as a different thing, it needs to be a different thing.
Do you have any plans to support Ultra HDR images, on the Android app and/or other platforms? I did some browsing of the site and didn't see any examples (but that doesn't mean it's unsupported).
I really enjoy that aspect for sharing photography on Instagram, and the support (at the client level, at least) is fairly straightforward (see https://developer.android.com/media/grow/ultra-hdr/display).
The idea of the platform does sound great, and the application seems very polished from what I can tell so far.
I actually had to look this up. Currently we don't but apparently we could support it if you upload via the web. Our mobile apps use Expo and the native image picker re-encodes the image.
If you're interested still despite limitations (Ultra HDR Images also won't be seen by many people) I can look into this more.
Thanks for taking a look! I haven't used Expo, but looking at the docs I wonder if using 'allowEditing: false' and 'exif: true' in the ImagePickerOptions would bypass reencoding? The gainmap is stored in the EXIF data. Instagram is similar in that the gainmap is stripped if you edit the image at all, cropping included, but if no crops or filters are done then the image retains the gainmap.
Either way if the clients are able to render the HDR images, I think uploading on web is still a usable implementation.
I just did some more research and we probably won't do the work until there's more demand. We can actually easily support it on desktop (we upload the raw bytes but always default to serving a compressed version, which I can override for HDR) but for mobile there's no drop-in solution for Expo apps and thus would have to build 4 new things (iOS and Android uploader and display components). That said, I'd like to do this because I want PIECES to have a rich visual experience and was thinking of recruiting more photographers, but probably not now. Thanks for your interest though.
That's fair enough, thanks for taking a look into it anyway!
Do you have some sort of detection for 'faceless accounts' ? Or are you moderating that by hand?
I like the idea, but how do you actually enforce the 'no bots' model?
Doing it by hand now. Besides the site itself, I have spent most of my time building out the backend automation. I imagine I can manage the first 10k users myself and opening it up slow enough that a community itself develops and will be self-policing.
I got back on instagram recently and the first thing I see when I login now is content from repost accounts and it's like, do I really want to see this? Is this what happens when we optimize for engagement blindly? A hypothesis I have is that if we make rules in the beginning this will be easier to enforce at scale. See an account that's just reposting stuff -> report it -> review
I hope this isn't coming off as snarky or anything, I am honestly interested in the project...
So you're thinking something like Twitter(X)'s community notes, but instead of flagging misinformation, it will be used to 'community flag' an account as AI? Maybe it shadow bans, maybe it brings it to moderation queue, whatever...
I've seen a few of these 'NO LLM ALLOWED' sites pop up now, but any countermeasure to LLM use is easily subverted. I think the only real way to keep the LLMs out is, as you seem to agree, some sort of user-reported system (But that seems open to fraud....difficult problem to solve!)
For sure. AI is only getting better so it's not a promise I can make but it's worth stating from the beginning. But the thing about Meta/Twitter/etc.,. is that they're financially incentivized to promote these accounts (more eyes = more ad revenue) so that makes it even worse. There will never be an explicit financial incentive to run bot accounts/repost farms on PIECES (I'm sure people will try though).
I recognize that we'll have to re-think things once we hit millions of users (I am an optimist). But for the first ten thousand or hundred thousand, I'm confident we can keep it a people-centric place and then strategize for future growth once our strategies start to be less effective.
Is this site inclusive of nonhuman identities? [0]
It seems incredibly focused on "real humans" etc, which is (common...), but my real identity doesn't exist in the physical world.
(There are also multiple of them thanks to DID, but that tends to require a significant rearchitecture to accommodate)
[0]: https://en.wikipedia.org/wiki/Otherkin
We'll have to open a committee to launch an investigation to produce a report about whether accounts held by Otherkin should be allowed.
Stay tuned and thanks for your patience.
FWIW, I think if you're serious about building, you probably don't have resources to allocate to committees on mental disorders.
That almost sounds satirical, but if you are serious about this, thank you. <3
Best of luck!
I’ve seen a few posts about a “better social network” tried a few even thought about making my own.
It seems hard to get them off the ground.
I wonder if the answer to the enshittification of social networks is “no social network at all”.
I'm going to sound a little arrogant here but part of it is because most people don't actually believe in what they're doing. I see posts of founders who write, "hey, we started a social network, but we've since pivoted to <insert AI tool here>".
I believe in PIECES to my core. Everything I'm doing now with my full-time job so I don't have to take outside investment and saving is to get not only PIECES but my dating platform Lettr Dating off the ground. PIECES will be around in 5 years, I guarantee it, because it's the social network that I always dreamed of having - even going back in the day to 2003 when I would just post on my xanga.
This doesn't seem to have been picked up by the HN algo, which is fine. But I'm going to keep at it.
If you're in NYC, I'll be promoting PIECES and Lettr in the parks at night time, and I'll likely be hiring an intern at NYU to help me spread it around campus. If you see this and are at NYU, I'd love to chat - https://piecesof.me/p/6qrjydihiq3tikg3fb85fqcelqnhph