Programmer and sysadmin (DevOps?), wannabe polymath in tech, science and the mind. Neurodivergent, disabled, burned out, and close to throwing in the towel, but still liking ponies 🦄 and sometimes willing to discuss stuff.

  • 4 Posts
  • 1.46K Comments
Joined 1 year ago
cake
Cake day: June 26th, 2023

help-circle



  • There is an allegation about him helping/inciting/collaborating/conspiring with Manning to break a password that would allow them to access information requiring a higher security clearance.

    It’s a serious accusation, and it’s compounded by suspicions of him favoring Russia in his filtering of leak releases, but it’s still crazy the amount of time he’s been not-free because of something he hasn’t been tried or found guilty of.


  • Yeah… “problem” was kind of tongue in cheek.

    But it’s not exactly a “default”, it’s more of a “demographic with little data”… and I bet it’s small enough that the algorithm is showing exactly the content most of its members are looking for. It’s somewhat of a sad reflection on the state of privacy, when keeping things private becomes a segmenting parameter.



  • Interesting article, but in my experience it overstates the problem… at least for Facebook itself (I have zero interaction with Instagram, Threads, or VR).

    I’ve gone back to Facebook for the last few months, and out of what it mentions, I’ve only seen like half of it, mostly in the comment sections.

    Or to be more precise, for 2024 Q2, I’m seeing:

    • election disinformation - almost none
    • violent content
    • child sexual abuse material
    • hate speech - only in comments
    • fake news - almost none
    • crypto scams - a few
    • phishing - a few
    • hacking
    • romance scams - almost none
    • AI content - almost none
    • uncanny valley stuff

    The article however forgot to include:

    • science deniers - a lot in open comments, very few in groups
    • religious zealots - in comments
    • political trolls - a few in comments
    • state-sponsored propagandists - a few in comments
    • general trolls - a few in comments

    Still interesting how I get close to zero of these in my main feed.

    there’s a level of disinvestment in Facebook

    Disagree. Facebook has reached a “plateau of stability” where the current moderation tools keep enough people on the platform to make it profitable.

    I’ve been actively reporting+blocking problematic content, and while about 99% of my reports end up in “no action was taken”, it works wonders to keep my feed and group comments clean.


  • Because I have basically nothing in my feed on this account, Facebook backfills it with “recommended” posts and I was pretty shocked at how universally terrible they are. […] since I’ve provided very little in the way of alternative information or interaction for it to use

    There is your problem.

    When an information-hungry platform like Google or Meta asks you to fill out your preferences “to serve you more relevant content”… they are not lying. I mean, it’s also to select ads that will pay more for your attention, but the thing with the content algorithm is, if you don’t give it data, then it will ass-u-me that you’re statistically most likely to engage with content that is getting most engaged… by people who have also not provided it any data.

    The problem with that cohort, is it not only includes the few people with legitimate security concerns, but also those who got dark secrets to hide, and/or are using “incognito” browser mode to look for porn.

    I don’t like to give too much info about myself, but I also don’t want to get stuff intended for the “average horny fanatics” group, so I try to give enough data for the algorithm to put me into a group that makes more sense to me.

    And it works. The strongest signal you can send to the algorithm, is blocking content you don’t want to see. It’s amazing how quickly modern algorithms learn to avoid showing me most porn, politics, or religious content, and instead show me science and humor. They still send like 1% of trash my way, clearly checking whether I’ll maybe engage with it, but report+block works wonders.


  • Be(e) nice cuts both ways: I don’t find the superiority implication of the capitalization to be nice, and OP’s explanations don’t make me think of anything nice behind them.

    It may boil down to something as simple as netiquette, where ALL CAPS MEANS SCREAMING, or AlTeRnAtInG cAsE means mocking… but the explanations seem to point more in the direction of asking to use MASTER/OWNER as someone’s pronouns. Not nice.


  • This is an interesting issue, with multiple fronts:

    • Spain has a large chunk of its GDP tied to tourism, but…
    • Some places are getting overcrowded, like Ibiza, Barcelona, or Madrid, which discourages tourism.
    • Having a large number or holiday apartments, increases housing prices for local residents…
    • And causes noise issues during the holiday season…
    • And makes it difficult for seasonal workers to find a place to live.
    • While during the off-season, it leaves a lot of apartments unoccupied, making them an easy target for illegal occupation, with whole gangs living from it, which then require LEO resources to vacate them…
    • And makes it extra difficult for local non-tourism businesses to survive…
    • To the point that they’re converting business locals into… holiday apartments.

    The plan to shut “all” holiday apartments is kind of a pipe dream, or part of populist politics… more so in Barcelona, where right now the recent elections have left Catalonia with parties so divided, that they can’t even agree on a viable candidate to lead it.

    A slightly more realistic issue to tackle, are “illegal” holiday apartments that don’t pay the corresponding taxes. Some estimate that Madrid has a 10:1 ratio of illegal vs. registered holiday apartments.

    But in general, there is currently no solution that would keep those apartments occupied all year round, without neighbor conflict, in areas that live mainly off seasonal tourism.

    For example, Ibiza has 40K permanent residents, but capacity for 600K tourists, which leads to seasonal workers living in trailer parks, or even in their own cars.








  • As usual, there is a grain of truth behind those claims:

    Cloudflare offers their own DNS, with stated benefits like filtering out some sites, while resolving CF sites… but some VPNs also set their own DNS, which don’t fully match Cloudflare’s… resulting in some combinations of CF site and VPN, not working. I’d blame the VPN for that, but people’s experience is going to be “everything works, except some CF sites” 🤷

    Cloudflare is a “potential” MITM: they claim not to read the traffic… but as a TLS terminator, they get the ability to read it without anyone’s knowledge.

    All non-encryped traffic is considered to be “insecure” for some time now. The whole point of initiatives like Let’s Encrypt, is to remove everyone on the client-server path from the list of entities you have to trust, so it ends up as: client software, client system, CertAuth, server owner, server software.

    Ideally, we’d have homomorphic encryption on the servers, but it’s not there yet.


  • From the first 15 min of the edited video: that FUTO boss is an embarrassment, good on Rossman to get him to change things.

    I don’t really want to watch the remaining hour, after someone says things like:

    • He didn’t follow the discussions back in the 2000s
    • OSI didn’t hijack the “open source” definition
    • Less than 1000 people would care
    • Asked his programmers, and they didn’t care

    I call BS. Weak excuses.

    There is a reason people say “FLOSS” instead of “Open Source”. There is a reason Stallman says what he says. There is a reason you can tell apart who understands what’s going on, by whether they understand the differences or not.


    A quick reminder:

    • Free - as in beer, not as in freedom
    • Libre - as in freedom
    • Open Source - you can see the source code

    Stallman created the GPL to allow people to see (open) and change (libre) the code (source)… then “pay forward” that freedom, in echange for being able to charge money (non-free) for their contributions.

    He often referred to it as simply “Open Source”… which turned out to be a mistake. Very soon (as in pre-1990), it became clear that there were two more competing camps for the “Open Source” definition:

    • Academia - people who got paid anyway, whether they saw a penny from their software or not
    • Business - who wanted to get as much money as possible, for as cheap as possible

    Both those camps aligned with licenses where developers gave up all their rights, but anyone could very easily take them back and claim as their own (“closing” the software). Famous examples are Microsoft, Apple, Google, Facebook, etc.

    The “Open Source Initiative” was created to gatekeep the “Open Source” definition, by keeping a list of licenses that were “OSI compliant”. A side effect of that gatekeeping, was erasing the understanding of the terms “Free” and “Libre” from the public’s minds.

    Plenty more than “1000 people” understood what was going on, and were against OSI, seeing it as an EEE move from the Business camp.

    People new to it, started using the term “open source” (as per OSI) without a care, only to later realize the Business camp was taking advantage of them… [surprised Pikachu face]


    This FUTO boss is not young or inexperienced, he’s a Business-man who, not surprisingly, decided to use a license with a closing clause, that he used the chance to call “Open Source” by exploiting people’s lack of understanding.


  • Let me clarify: I’ve seen the sand bucket guy’s art featured twice on the news in the past few days, filmed at an art gallery, described as art, commented as being art. It’s not some random event, it’s the current publicly accepted definition of “art”.

    My statement, not insinuation, as to why AI art is comparable to “traditional” art, comes after that.

    What comes across as desperate however, is generalizing all AI output and disparaging it, without considering the quality of input from the person behind it. Reminds me of how photography used to not be art, how electric instruments couldn’t be art, or how using a computer couldn’t be art either. Tools don’t make or break an artist.