Decentralized Social Media Is the Only Alternative to the Tech Oligarchy
Decentralized Social Media Is the Only Alternative to the Tech Oligarchy
The TikTok ban and Donald Trump's rise to power show how fragile our social media accounts are. We must normalize and invest in decentralized social media.
You're viewing a single thread.
Agreed. But we need a solution against bots just as much. There's no way the majority of comments in the near future won't just be LLMs.
Decentralized authentication system that support pseudonymous handles. The authentication system would have optional verification levels.
So I wouldn't know who you are but I would know that you have verified against some form of id.
The next step would then by attributes one of which is your real name but also country of birth, race, gender, and other non-mutable attributes that can be used but not polled.
So I could post that I am Bob living in Arizona and I was born in Nepal and those would be tagged as verified, but someone couldn't reverse that and request if I want to post without revealing those bits of data.
we have to use trust from real life. it's the only thing that centralized entities can't fake
I feel like it's only a matter of time before most people just have AI's write their posts.
The rest of us with brains, that don't post our status as if the entire world cares, will likely be here, or some place similar... Screaming into the wind.
I feel like it’s only a matter of time before most people just have AI’s write their posts.
That's going right into /dev/null as soon as I detect it-- both user and content.
Instances that don’t vet users sufficiently get defederated for spam. Users then leave for instances that don’t get blocked. If instances are too heavy handed in their moderation then users leave those instances for more open ones and the market of the fediverse will balance itself out to what the users want.
I wish this was the case but the average user is uninformed and can’t be bothered leaving.
Otherwise the bigger service would be lemmy, not reddit.
the market of the fediverse will balance itself out to what the users want.
Just like classical macroeconomics, you make the deadly (false) assumption that users are rational and will make the choice that’s best for them.
The sad truth is that when Reddit blocked 3rd party apps, and the mods revolted, Reddit was able to drive away the most nerdy users and the disloyal moderators. And this made Reddit a more mainstream place that even my sister and her friends know about now.
We could ask for anonymous digital certificates. It works this way.
Many countries already emit digital certificates for it's citizens. Only one certificate by id. Then anonymous certificates could be made. The anonymous certificate contains enough information to be verificable as valid but not enough to identify the user. Websites could ask for an anonymous certificate for register/login. With the certificate they would validate that it's an human being while keeping that human being anonymous. The only leaked data would probably be the country of origin as these certificates tend to be authentificated by a national AC.
The only problem I see in this is international adoption outside fully developed countries: many countries not being able to provide this for their citizens, having lower security standards so fraudulent certificates could be made, or a big enough poor population that would gladly sell their certificate for bot farms.
Your last sentence highlights the problem. I can have a bot that posts for me. Also, if an authority is in charge of issuing the certificates then they have an incentive to create some fake ones.
Bots are vastly more useful as the ratio of bots to humans drops.
Also the problem of relying on a nation state to allow these certificates to be issued in the first place. A repressive regime could simply refuse to give its citizens a certificate, which would effectively block them from access to a platform that required them.
I mentioned this in another comment, but we need to somehow move away from free form text. So here’s a super flawed makes-you-think idea to start the conversation:
Suppose you had an alternative kind of Lemmy instance where every post has to include both the post like normal and a “Simple English” summary of your own post. (Like, using only the “ten hundred most common words” Simple English) If your summary doesn’t match your text, that’s bannable. (It’s a hypothetical, just go with me on this.)
Now you have simple text you can search against, use automated moderation tools on, and run scripts against. If there’s a debate, code can follow the conversation and intervene if someone is being dishonest. If lots of users are saying the same thing, their statements can be merged to avoid duplicate effort. If someone is breaking the rules, rule enforcement can be automated.
Ok so obviously this idea as written can never work. (Though I love the idea of brand new users only being allowed to post in Simple English until they are allow-listed, to avoid spam, but that’s a different thing.) But the essence and meaning of a post can be represented in some way. Analyze things automatically with an LLM, make people diagram their sentences like English class, I don’t know.
A bot can do that and do it at scale.
I think we are going to need to reconceptualize the Internet and why we are on here at all.
It already is practically impossible to stop bots and I'm a very short time it'll be completely impossible.
I think I communicated part of this badly. My intent was to address “what is this speech?” classification, to make moderation scale better. I might have misunderstood you but I think you’re talking about a “who is speaking?” problem. That would be solved by something different.
It sounds like you're describing doublespeak from 1984.
Simplifying language removes nuance. If you make moderation decisions based on the simple English vs. what the person is actually saying, then you're policing the simple English more than the nuanced take.
I've got a knee-jerk reaction against simplifying language past the point of clarity, and especially automated tools trying to understand it.
What? I post a lot, but the majority?
...oh, you said LLM. I thought you said LMM.
A simple thing that may help a lot is for all new accounts to be flagged as bots, requiring opt out of the status for normal users. It's a small thing, but any barrier is one more step a bot farm has to overcome.
Reputation systems. There is tech that solves this but Lemmy won’t like it (blockchain)
You don't need blockchain for reputations systems, lol. Stuff like Gnutella and PGP web-of-trust have been around forever. Admittedly, the blockchain can add barriers for some attacks; mainly sybil attacks, but a friend-of-a-friend/WoT network structure can mitigate that somewhat too,
Slashdot had this 20 years ago. So you're right this is not new.or needing some new technology.
Space is much more developed. Would need ever improving dynamic proof of personhood tests
I think a web-of-trust-like network could still work pretty well where everyone keeps their own view of the network and their own view of reputation scores. I.e. don't friend people you don't know; unfriend people who you think are bots, or people who friend bots, or just people you don't like. Just looked it up, and wikipedia calls these kinds of mitigation techniques "Social Trust Graphs" https://en.wikipedia.org/wiki/Sybil_attack#Social_trust_graphs . Retroshare kinda uses this model (but I think reputation is just a hard binary, and not reputation scores).
I dont see how that stops bots really. We're post-Turing test. In fact they could even scan previous reputation points allocation there and divise a winning strategy pretty easily.
I mean, don't friend, or put high trust on people you don't know is pretty strong. Due to the "six degrees of separation" phenomenon, it scales pretty easily as well. If you have stupid friends that friend bots you can cut them off all, or just lower your trust in them.
"Post-turing" is pretty strong. People who've spent much time interacting with LLMs can easily spot them. For whatever reason, they all seem to have similar styles of writing.
I mean, don’t friend, or put high trust on people you don’t know is pretty strong. Due to the “six degrees of separation” phenomenon, it scales pretty easily as well. If you have stupid friends that friend bots you can cut them off all, or just lower your trust in them.
Know IRL? Seems it would inherently limit discoverability and openness. New users or those outside the immediate social graph would face significant barriers to entry and still vulnerable to manipulation, such as bots infiltrating through unsuspecting friends or malicious actors leveraging connections to gain credibility.
“Post-turing” is pretty strong. People who’ve spent much time interacting with LLMs can easily spot them. For whatever reason, they all seem to have similar styles of writing.
Not the good ones, many conversations online are fleeting. Those tell-tale signs can be removed with the right prompt and context. We're post turing in the sense that in most interactions online people wouldn't be able to tell they were speaking to a bot, especially if they weren't looking - which most aren't.
Do you have a proof of concept that works?
Are they just putting everything on layer 1, and committing to low fees? If so, then it won't remain decentralized once the blocks are so big that only businesses can download them.
It has adjustable block size and computational cost limits through miner voting, NiPoPoWs enable efficient light clients. Storage Rent cleans up old boxes every four years. Pruned (full) node using a UTXO Set Snapshot is already possible.
Plus you don't need to bloat the L1, can be done off-chain and authenticated on-chain using highly efficient authenticated data structures.
Also is data scraping as much of an issue?
Data scraping is a logical consequence of being an open protocol, and as such I don't think it's worth investing much time in resisting it so long as it's not impacting instance health. At least while the user experience and basic federation issues are still extant.
There are simple tests to out LLMs, mostly things that will trip up the tokenizers or sampling algorithms (with character counting being the most famous example). I know people hate captchas, but it’s a small price to pay.
Also, while no one really wants to hear this, locally hosted "automod" LLMs could help seek out spam too. Or maybe even a Kobold Hoard type "swarm."
Captchas don't do shit and have actually been training for computer vision for probably over a decade at this point.
Also: Any "simple test" is fixed in the next version. It is similar to how people still insist "AI can't do feet" (much like rob liefeld). That was fixed pretty quick it is just that much of the freeware out there is using very outdated models.
I'm talking text only, and there are some fundamental limitations in the way current and near future LLMs handle certain questions. They don't "see" characters in inputs, they see words which get tokenized to their own internal vocabulary, hence any questions along the lines of "How many Ms are in Lemmy" is challenging even for advanced, fine tuned models. It’s honestly way better than image captchas.
They can also be tripped up if you simulate a repetition loop. They will either give a incorrect answer to try and continue the loop, or if their sampling is overturned, give incorrect answers avoiding instances where the loop is the correct answer.
They don’t “see” characters in inputs, they see words which get tokenized to their own internal vocabulary, hence any questions along the lines of “How many Ms are in Lemmy” is challenging even for advanced, fine tuned models.
And that is solved just by keeping a non-processed version of the query (or one passed through a different grammar to preserve character counts and typos). It is not a priority because there are no meaningful queries where that matters other than a "gotcha" but you can be sure that will be bolted on if it becomes a problem.
Again, anything this trivial is just a case of a poor training set or an easily bolted on "fix" for something that didn't have any commercial value outside of getting past simple filters.
Sort of like how we saw captchas go from "type the third letter in the word 'poop'" to nigh unreadable color blindness tests to just processing computer vision for "self driving" cars.
They can also be tripped up if you simulate a repetition loop.
If you make someone answer multiple questions just to shitpost they are going to go elsewhere. People are terrified of lemmy because there are different instances for crying out loud.
You are also giving people WAY more credit than they deserve.
@NuXCOM_90Percent @brucethemoose would some kind of proof of work help solve this? Ifaik its workingnon tor
Somehow I didn’t get pinged for this?
Anyway proof of work scales horrendously, and spammers will always beat out legitimate users of that even holds. I think Tor is a different situation, where the financial incentives are aligned differently.
But this is not my area of expertise.
Well, that's kind of intuitively true in perpetuity
An effective gate for AI becomes a focus of optimisation
Any effective gate with a motivation to pass will become ineffective after a time, on some level it's ultimately the classic "gotta be right every time Vs gotta be right once" dichotomy—certainty doesn't exist.
We also need a solution to fucking despot mods and admins deleting comments and posts left-and-right because it doesn't align with their personal views.
I've seen it happen to me personally across multiple Lemmy domains (I'm a moron and don't care much to have empathy in my writing, and it sets these limp-wrist morbidly obese mods/admins to delete my shit and ban me), and it happens to many people as well.
Don’t go blaming your inability to have empathy on adhd. That is in absolutely no way connected. You’re just a rude person.
I'm also rude in real life too! 😄
You have that tool, it’s called finding or hosting your own instance.
lemm.ee and lemmy.dbzer0.com both seem like very level-headed instances. You can say stuff even if the admins disagree with it, and it's not a crisis.
Some of the big other ones seem some other way, yes.
Communities should be self moderated. Once we have that we can really push things forward.
Self Moderated is just fine. Why do I need to doxx myself to be online? I'm not giving away my birth certificate or SSN just to post on social media that idea is crazy lmao.