Following on my from my recent post on free speech at X/Twitter that exposed how poor X is at free speech and how the Hive blockchain is hugely superior.. Here's an update in response to Musk's latest brainwave - no X money for you if people disagree with you!
I have continued to follow the saga of Elon Musk's adjustments to X with interest, especially considering that he is tackling the problems that we at Hive have already mostly tackled. In particular, he advertised himself and the X network as being a place for free speech, which would effectively 'save the world' - this is Hive's back yard. Today he has announced yet another step down the continual slippery slope of censorship by announcing that monetisation on X will be revoked for posts that have 'community notes' written for them. What does this mean?
https://twitter.com/ura_soul/status/1718677407052247390
- ura_soul
Remember back when Musk was claiming that Twitter had been totally blackmailed and controlled by the ADF for years? They were pressuring advertisers to stop advertising if Twitter didn't censor posts for them. We'd better hope that community notes is a robust solution for discerning both genuine community intent and the truth or he's just created a system that makes the ADFs job even easier for them.
Free Speech
In my experience, there really are only two meaningful positions when it comes to being a 'free speech advocate':
You support absolute free speech, which means you personally allow for all people, everywhere, to say whatever they want, whenever they want - without fear of being punished for it.
You like the idea of free speech but are troubled by what some people say and you want to control some people, eliminating their free speech in some situations.
Option 1, absolute free speech, is - to me - the only real free speech. Option 2 is an absurd caricature of free speech, which amounts to virtue signalling at best or evil at worst. This is because there is no practical way to censor 'some' people 'some of the time' without the situation becoming a power struggle of opinions and individual egos/wills. Who is to say which view or person is wrong/dangerous? Typically, it is the most powerful person/group who ultimately gets to decide this - which is a situation that comes with absolutely no guarantee that their decision will be correct or even safe.
Elon Musk has claimed in the past that he is a 'free speech absolutist', but this was clearly untrue, since everything he has done at X has been based on option 2 and not option 1. The websites that run on the Hive blockchain come far closer to free speech absolutism than X/Twitter ever has done or is likely to.
So given that we know that Elon falls into the group of people who want to control free speech while virtue signalling that they are super pro 'definitely bro' supportive of free speech.. Where does this leave us?
I have already written on why the monetisation at X is deceiving and terrible compared to Hive, so all that remains is to explore this issue of demonetisation of tweets that have attracted a 'community note'.
This new 'feature' apparently makes sense to some people, who are tired of seeing 'online celebrities' getting rich from peddling lies. That's understandable, but they are short sighted here. There's no way to prevent this without also stifling the behaviour/expression/growth of accounts that actually do stick to the truth. Do you honestly think that a world that enables liars to get paid due to community viewership (support) won't also use the control systems put in place to similarly warp the flow of information in the X community?
"Well, Community notes is transparent and not controlled by Elon Musk"
Oh ok, cool - that's perfect then - a flawless system.. The first ever created, nice! ;)
Forgive my cynicism but I am old enough to have seen the evolution of Wikipedia - a system that was claimed to be similar to X's community notes - a community driven system of discerning truth. The problem is that (as even Elon Musk has publicly agreed), Wikipedia has long been entirely corrupted behind the scenes. You only need to experiment by making an update to a page on Wikipedia to any topic that some group somewhere might have a vested interest in - to discover how quickly your comment is rejected/overwritten and how the people involved are not even remotely interested in actual truth.
Digging into community notes?
So how well does community notes do at this challenging problem of throwing a question at 'the community' and said community then spitting out 'absolute truth' (or close to it)? This is the core of the problem because if the X community is biased, ignorant, in denial and/or bribed then the demonetisation of posts based on community notes is a recipe for disaster - at least from the perspective of X being the useful tool for finding 'facts' that Musk says he wants it to become.
My own investigation into this topic was somewhat thwarted at the first hurdle - I can't join the community notes group because my phone isn't running on a 'trusted network'. There is no list of trusted networks, so I can't tell you why that is! The fact that my phone's network is one of the biggest in England seems to be irrelevant. 'Sorry, you can't join our truthy club' because of 'reasons'.
Not a good start. (see the page on this at X here).
Ok, so the only other option I have for investigating this is to look at the source code for the algorithm that is used to manage the community notes feature, since it's public in Github here. Note: The page on X that is meant to point to this Repo, actually doesn't - I had to find it in a search engine.
I have a couple of problems with this source code atm, which hopefully people on Hive can help me with:
- I don't use Python, which is the language used to write the code.
- I am exceptionally busy with other projects and don't have a lot of time throw at analysing their code atm anyway.
Maybe someone on Hive that knows python has time to go through the source code and to write an appraisal of it, highlighting flaws/risks etc?
Please do if you can!
Diversity!
But wait, there's a page on X that explains how they ensure that community notes are reliable - this one on diversity of perspectives
This is the only page I've so far found that gives any description of why anyone should trust Community notes, so it's kind of important. Here's the full page as it stands:
Community Notes aims to identify notes that many people on X will find helpful, including people with different points of view.
To find notes that are helpful to the broadest possible set of people, Community Notes takes into account not only how many contributors rated a note as helpful or unhelpful, but also whether people who rated it seem to come from different perspectives.
Community Notes assesses "different perspectives" entirely based on how people have rated notes in the past; Community Notes does not ask about or use any other information to do this (e.g. demographics like location, gender, or political affiliation, or data from X such as follows or posts). This is based on the intuition that Contributors who tend to rate the same notes similarly are likely to have more similar perspectives while contributors who rate notes differently are likely to have different perspectives. If people who typically disagree in their ratings agree that a given note is helpful, it's probably a good indicator the note is helpful to people from different points of view.
This approach has a number of benefits. First, it reflects the reality that people’s views can be nuanced, rather than defined by demographics. Second, in support of our focus on transparency, it allows people working with Community Notes public data to replicate, analyze and audit how Community Notes works, as it allows Community Notes to run entirely on publicly available data.
We are constantly evaluating ways to improve this approach (and welcome suggestions) This current method has shown promising results in helping find quality notes: in surveys of people who use X in the US, the majority of respondents found notes that earned a status of "Helpful" by Community Notes contributors to be “somewhat” or “extremely” helpful — this includes people from across the political spectrum.
In addition to the approach described here, we also work to understand specifically how helpful notes are to people from different political perspectives. For example, by analyzing X's follow, like, and repost graphs. It's important to note that these analyses don't directly impact the ratings of specific notes. Instead, they serve as a quality measure that links to specific notes and aids in refining our open-source algorithms. Learn more here.
So in summary, they find people who 'usually disagree' and if they agree on an idea then the idea is taken to be 'probably true'. Uhm.. Well, that's not really how truth works at all is it? That might be a good enough way to filter out a certain perspective of truth from fiction, but it's never going to even remotely come close to being reliable. Could it be reliable enough to justify demonitisation? No! lol.
Transparent algorithms seem great on the surface, but they always have a weak point - they can be studied and gamed. The only possible way to prevent people gaming this algorithm is to rigidly enforce KYC type identities, where every user is known and there are no bots. X does not have this environment and the presence of bots is abundantly clear. With the rise of AI systems it should be abundantly clear that gaming community notes using AI bots might actually be relatively straight forward.
In the middle of all that noise, is there a way for Elon Musk himself to manipulate the outcome of the algorithm in such a way? Of course - he even has super advanced AI! lol.
Summation
I do know that I have seen comments thrown up in X within community notes that amount to nothing more than unsubstantiated opinion - so I am unclear exactly how the community notes amount to much more than just another reply posted under the original tweet that has been boosted to be the 'official community position' (that you weren't part of deciding on and which probably doesn't reflect or consider all possible views/evidence).
The algorithm X is using to manage the notes is likely not that hard to game.
All of this means that the already hard to achieve monetisation on X just got even harder to reach for most people and even more possible to manipulate for those with the resources to do so.
Please don't rely on X as a source of truth alone. Always question everything and develop your own intellect, it's part of why you are alive on Earth in the first place!
Wishing you well,
Ura Soul
Powerful insights into the Hive blockchain are available at my website, Hive Alive.
Including the only way to track downvotes on Hive - The Untrending report
