sapient [they/them]

Autistic queer trans²humanist and anarchist. Big fan of dense cities, code, automation, neurodiversity, and self-organising resilient networks.

Pronouns: they/them, xe/xem, ze/zem

Favourite Programming Language: Rust

Alt-Account Of: @sapient_cogbag@sh.itjust.works

  • 3 Posts
  • 20 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle






  • VeilID might be something you find interesting. It’s designed to solve exactly this problem by enabling most nodes to NATsmash with help for p2p stuff, and also provides a general and very strong privacy framework including torlike routing .

    It was only unveiled at defcon this year though so the team behind it (Cult Of The Dead Cow) are trying to put docs in place ;p

    Its completely written in rust, easily embeddable, has good content locality and is probably the cleanest, most performant, and most easily integrated into projects architecture for stuff like this that I’ve seen, as a programmer who’s into this space and familiar with things like i2p, tor, etc. I really hope this one takes off, and the quality of it means I really think it could (at least once they throw the docs together ;p)




  • Be real. The cost of building means they’re always going to favour the wealthy. At best right now were running public copies of the older and smaller models. Local AI will always be running behind the state of the art big proprietary models, which will always be in the hands of the richest moguls and companies in the world.

    Distribution of LoRA-style fine-tuning weights means that FOSS AI systems have a long term advantage because of compounding effects. .

    That is, high-quality data provided for smaller models and very small “model finetuning” weights, which is more accessible to open groups, are sufficiently accessible and modular in their improvements to a given model that the FOSS community can take and run with it to compete effectively with proprietary groups from even a single leak.

    Furthermore, smaller and more efficient models which can be run on lower end hardware also avoid the need to send off potentially sensitive data to AI companies and enable the kinds of FOSS compounding effect explained above.

    This doesn’t just affect people who like privacy, but also companies with data privacy requirements . - as long as the medium models are “good enough” (which I think they are ;p), the compounding effects of LoRA tuning and better data privacy properties, and further developments which already exist in research papers towards much lower weight-count models and training mechanisms capable of greater weight efficiency to induce zero-shot learning, mean local AI can compete with proprietary stuff. It’s still early days but it is absolutely doable even today with fairly low-end hardware, and it can only get better for the reasons provided.

    Furthermore, “intellectual property” and copyright stuff have an absolutely massive and arguably even more powerful set of industries behind them. Trying to strengthen IP stuff against AI means that AI will only be available to those controlling these existing IP resources and it’s unending stranglehold on technology and communication and people as a whole :/

    AI I think is also forcing more and more people to look and reevaluate society’s relationship with work and labour. And frankly I think that this is super important, as it enables a greater chance of more radical liberation from the existing structures of not just capitalism and it’s hierarchies but the near-mandatoriness of work as a whole (though there has already been some stuff like this around the concepts of “bullshit jobs”).

    I think people should use this as an opportunity to unionise and also try and push for cooperative and democratic control of orgs ., and many other things that I CBA to list out ;3









  • Profiling happens on the server end, but the unique identification can happen either on the server end or the client end or both.

    And it’s Meta. Their entire organisation is dedicated to manipulation, data collection, etc. - hell, they might do the profiling inside the app and only send results. What’s “realistic” is expecting them to spy on you, not giving an organisation like them any benefit of the doubt!



  • The post is too big for my next edit, so here is the next edit in a comment:

    Edit 2 - Clarification, Expanding on Facebook’s Behaviour, Discussion of Admin-FB Meetups

    I want to clarify the specific dangers of Meta/FB, as well as some terminology.

    Embrace, Extend, Extinguish, and Embrace, Extend, Consume

    The link I posted approximately explains EEE, but in this thread I’ve used the phrase “Embrace, Extend, Consume”, to illustrate a slightly modified form of this behaviour.

    Embrace, Extend, Consume is like Embrace, Extend, Extinguish except the end goal isn’t complete annihilation of the target. Instead of defederating at the endpoint, Meta/FB just dominates the entire standard, and anyone who steps out of line is forced into a miniscule network of others.

    They can then use this dominant position to buy out or consume large instances, or for example, force data collection features into the standard and aggressively defederate anyone else who doesn’t comply >.< - because they’re so big, most instances will comply in the service of “content”.

    Such a dominant position can even be obtained simply by sheer user mass, which Threads already has to some degree, as long as the relevant instance has large amounts of financial resources to buy out instances.

    In this way, they consume the network entirely, which doesn’t necessarily destroy the communities but essentially Borg-ifies them and renders people unable to leave their grasp.

    Facebook/Meta-Specific Threats: Information Warfare & Manipulation

    One of the major specific threats of Meta/FB in particular is their long and continued history of engaging in what essentially amounts to large-scale psychological manipulation and information warfare towards it’s various goals (money, total domination of human communication, subsuming the internet in countries where the infrastructure is still too small to resist a single corporation restricting it’s content, political manipulation, collection of ever more data, etc.), against both it’s users and non-users.

    They have well over a decade of experience in this, hundreds of times more users than us (providing good cloaking for astroturfers), and untold amounts of labour, research and other resources have been poured specifically into figuring out the most effective ways to manipulate social groups via techniques like astroturfing, algorithmic prioritization, and more sophisticated strategies I am not aware of. All backed by data from literally billions of human beings >.<

    This means that exposing the Fediverse to Facebook/Meta is essentially exposing us all to one of the most organised and sophisticated information warfare machines that has ever been created. Cutting off the connections immediately (as in the other analogy by @BreakingBad@lemmy.world) not only protects from direct EEE/EEC, but also makes it harder for Meta/Facebook to influence, dominate, and consume the conversation here, either by sheer user-mass, or by malicious information warfare (or even unintentional consequences of their algorithms), or by a combination of all of these.

    We know they are extremely malicious and willing to use these methods towards real-life, ultra-harmful ends. Examples are at the start of this post :)

    For hypothetical examples on how this might work - in reality it might be different in the specifics (these are just illustrative):

    • Meta/FB could start a campaign (maybe astroturfed) for “user safety”, where they encourage people to distrust users from smaller instances or any user with their instance address marker not on @threads.<whatever their url>
    • Meta/FB could add “secure messaging” (lol, it’s facebook), but only between threads users. Then they could push the idea that ActivityPub is bad for privacy (the DMs are, but just use Matrix ;p - if you post stuff publicly, it makes sense that it’s public).
    • Meta/FB could by simple user mass result in most communities being on Threads. People tend to drift towards more populous communities about the same topic, in general, and Threads unbalances the user ratios so much that everyone would just go to those >.< (as opposed to right now, where we have similar sized communities on several large instances, where most people subscribe to most of them)
    • Meta/FB could use social engineering to push for changes to the ActivityPub protocol that are harder for other ActivityPub servers to implement ^.^, or even ones that are hard for non-proprietary clients to implement. For example, embedding DRM in the protocol or something like that.
    • Meta’s algorithms could over time shift towards deprioritising non-“paid”/“verified” Threads users.
    • It’s already been explained how the app as we know it essentially makes it hard for people to leave due to the fact only they have access to their server software and they also ensure that the app is only a specific client for this service.

    Instance Admins, and the “Friendliness” of Meta

    Some instance admins have been in contact with Meta/FB. It does make sense for at least some of them to do “due dilligence”, but I’ve seen in at least one post a comment on the friendliness and cooperativeness of the engineers and the fact they mostly discussed architectural concerns and stuff like moderation and technical stuff.

    I want to remind instance admins that no matter how nice the engineers are - and how much they share your interests - they are still working for what is essentially a mass information warfare machine. This doesn’t make them malicious at all, but it does mean that what they are doing is not a solid perspective on the actual goals and attitude of Meta/Facebook, The Corporate Assimilator Organism.

    Regardless of what they have discussed, they are obligated as employees to act on Meta’s orders, not the things they actually want to work on or the things both them and you find important ^.^ - or even act towards the goals they want to act towards when Meta inevitably goes for the throat.

    I encourage instance admins to keep this in mind, and further keep in mind that Meta is pretty much royalty when it comes to social stuff and how to appeal to people. If they were trying to appeal to a more corporate social media service, they’d probably have gone with sending in the C-suite, but they know this community is technically inclined and less likely to buy into corpo speak and corpo bullcrap, so they probably hooked you up with all the chill engineers instead :).

    Reiterating my view: Resist Corpo-Assimilation!

    Note on This Post

    I’ve realised this post would probably be most useful if the primary targets of Threads could see it (Mastodon). But I don’t have Mastodon cus I really am not into microblogging myself, so RIP ;p