I like to code, garden and tinker
“From March 1, 2024, an order will come into force to block VPN services providing access to sites banned in Russia,” Sheikin was quoted as saying by state news agency RIA.
I assume this means it’s regarding outgoing communications, for censorship purposes most likely. I’d be surprised if they were blocking incoming VPN traffic, and I don’t think the Russian government has an issue with Yandex operating.
TL;DR: The bot is configured to condense certain instances and communities. At the moment, only beehaw.org is marked to be condensed.
Quickly looking at the source code, it seems ReplyToPostsCommand uses a SummaryTextWrapper, which contains an iterable for both CondensedSummaryTextWrapperProvider and DefaultSummaryTextWrapperProvider. The DefaultSummaryTextWrapperProvider has a priority of -1_000 (so it’s always checked last) and is set to always return true on the supports(Community $community): bool. CondensedSummaryTextWrapperProvider references the config/services.yaml for it’s supports(Community $community): bool call which lists 0 condensed communities and 1 condensed instance, being beehaw.org.
If you are expecting a more windows-like experience, I would suggest using Ubuntu or Kubuntu (or any other distro using Gnome/KDE), as these are much closer to a modern Windows GUI. With Ubuntu, I can use the default file manager (nautilus) and do Ctrl+F and filter files via *.ext, then select these files then cut and paste to a new folder (drag and drop does not seem to work from the search results). In Kubuntu, the search doesn’t recognize * as a wildcard in KDE’s file manager (dolphin) but does support drag/drop between windows.
You can add streamlink-ttvlol with one of the known compatible proxies to remove the ads. Works great for me.
Also if there are issues with VLC crashing, I recommend MPV as it handles malformed stream data a little better. VLC will work great 99% of the time on twitch though.
Also, any plugin that Twitch doesn’t like (for example TTV LOL) is detected and will prevent a log in. You’ll need to disable the plugin to login, but can use it after logging in.
Thermometers, like most measurement devices, are always accurate until you get two of them. Each device has a specific tolerance (or should, otherwise it’s probably a horrible tolerance), for a grill thermometer this will look like -/+5C/10F. Additionally, everything used to read a measurement needs to be calibrated regularly to ensure proper function, otherwise readings cannot be trusted. For a thermometer, the easily accessible way to calibrate are to use ice water (does it read 0C/32F) and boiling water (does it read 100C/212F). Using these constants will allow you to adjust your thermometer and get a (more) accurate reading.
I also fail to see how this applies here. What is the disinformation? Where is the Russian bias? If you are seeing something I am not, please elaborate, but the summary in the article is:
No one would blame Zelenskyy for choosing the lesser of two evils here: Western banks over Russian tanks. Yet, the grim fact remains that even if his nation succeeds in repealing the Russian invasion, the future in store for Ukraine is not necessarily one of sovereignty and self-determination but, most likely, one of Western economic tutelage.
Of course large global asset managers are going to see money signs in their eyes. The fact is that Ukrainians are being put between a rock and a hard place, and exploitation of those kind of situations is capitalism 101.
Also, if you are assuming this is Russian propaganda, why is it coming from a website ran by a British political activist funded by a British investor. It also seems to be “mostly factual”. I’m failing to see where the tie to Russia is.
In my humble opinion, we too are simply prediction machines. The main difference is how efficient our brains are at the large number of tasks given for it to accomplish for it’s size and energy requirements. No matter how complex the network is it is still a mapped outcome, just the number of factors weighed is extremely large and therefore gives a more intelligent response. You can see this with each increment in GPT models that use larger and larger parameter sets giving more and more intelligent answers. The fact we call these “hallucinations” shows how effective the predictive math is, and mimics humans abilities to just make things up on the fly when we don’t have a solid knowledge base to back it up.
I do like this quote from the linked paper:
As we will discuss, we find interesting evidence that simple sequence prediction can lead to the formation of a world model.
That is to say, you don’t need complex solutions to map complex problems, you just need to have learned how you got there. It’s never purely random attempts at the problem, it’s always predictive attempts that try to map the expected outcomes and learn by getting it right and wrong.
At this point, it seems fair to conclude the crow is relying on more than surface statistics. It evidently has formed a model of the game it has been hearing about, one that humans can understand and even use to steer the crow’s behavior.
Which is to say that it has a predictive model based on previous games. This does not mean it must rigidly follow previous games, but that by playing many games it can see how each move affects the next. This is a simpler example because most board games are simpler than language with less possible outcomes. This isn’t to say that the crow is now a grand master at the game, but it has the reasoning to understand possible next moves, knows illegal moves, and knows to take the most advantageous move based on it’s current model. This is all predictive in nature, with “illegal” moves being assigned very low probability based on the learned behavior the moves never happen. This also allows possible unknown moves that a different model wouldn’t consider, but overall provides what is statistically the best move based on it’s model. This allows the crow to be placed into unknown situations, and give an intelligent response instead of just going “I don’t know this state, I’ll do something random”. This does not always mean this prediction is correct, but it will most likely be a valid and more than not statistically valid move.
Overall, we aren’t totally sure what “intelligence” is, we are just an organism that has developed more and more capabilities to process information based on a need to survive. But getting down to it, we know neurons take inputs and give outputs based on what it perceives is the best response for the given input, and when enough of these are added together we get “intelligence”. In my opinion it’s still all predictive, its how the networks are trained and gain meaning from the data that isn’t always obvious. It’s only when you blindly accept any answer as correct that you run into these issues we’ve seen with ChatGPT.
Thank you for sharing the article, it was an interesting article and helped clarify my understanding of the topic.
Disclaimer: I am not an AI researcher and just have an interest in AI. Everything I say is probably jibberish, and just my amateur understanding of the AI models used today.
It seems these LLM’s use a clever trick in probability to give words meaning via statistic probabilities on their usage. So any result is just a statistical chance that those words will work well with each other. The number of indexes used to index “tokens” (in this case words), along with the number of layers in the AI model used to correlate usage of these tokens, seems to drastically increase the “intelligence” of these responses. This doesn’t seem able to overcome unknown circumstances, but does what AI does and relies on probability to answer the question. So in those cases, the next closest thing from the training data is substituted and considered “good enough”. I would think some confidence variable is what is truly needed for the current LLMs, as they seem capable of giving meaningful responses but give a “hallucinated” response when not enough data is available to answer the question.
Overall, I would guess this is a limitation in the LLMs ability to map words to meaning. Imagine reading everything ever written, you’d probably be able to make intelligent responses to most questions. Now imagine you were asked something that you never read, but were expected to respond with an answer. This is what I personally feel these “hallucinations” are, or imo best approximations of the LLMs are. You can only answer what you know reliably, otherwise you are just guessing.
Best I could find is here, which is an article by Randall Munroe (the xkcd artist), and states:
davean (the xkcd sysadmin) wrote the patch
This blog post links to another wayback machine page (thank you archive.org!) here, which explains the sorting algorithm and states it’s original author:
Fortunately, the math for this was worked out in 1927 by Edwin B. Wilson.
Sounds like some QoS software is also limiting LAN traffic, seeing as it still works if the internet is disconnected. I would look if your router has “Adaptive QoS” or something similar enabled.
Does the flash drive show when you run lsblk with the correct amount of space? dd will overwrite the partition table and works directly with the underlying physical blocks of the device. If the flash drive isn’t broken, you should be able to rebuild the partition table with parted (tutorial from linuxconfig.org on the matter)
In most setups I have seen, the nginx instance provided by Lemmy is used due to the routing needed between lemmy/lemmy-ui being handled in nginx. Your reverse proxy can then point to the nginx instance to expose lemmy.
As for the data transfer costs, any network data originating from AWS that hits an external network (an end user or another region) typically will incur a charge. To quote their blog post:
A general rule of thumb is that all traffic originating from the internet into AWS enters for free, but traffic exiting AWS is chargeable outside of the free tier—typically in the $0.08–$0.12 range per GB, though some response traffic egress can be free. The free tier provides 100GB of free data transfer out per month as of December 1, 2021.
So you won’t be charged for incoming federated content, but serving content to the end user will count as traffic exiting AWS. I am not sure of your exact setup (AWS pricing is complex) but typically this is charged. This is probably negligible for a single-user instance, but I would be careful serving images from your instance to popular instances as this could incur unexpected costs.
As for the article, I think this is generally PR and corporate speak. Whatever their reasons were, they apparently didn’t shut down the initial XMPP servers until 2022 so it was a reliable technology. There “simplification” was bringing users into their ecosystem to more easily monetize their behaviour. This goes along with your last paragraph, at the end of the day the corporation is a for-profit organization. We can’t trust a for-profit organization to have the best of intentions, some manager is aiming to meet a metric that gets them their bonus. Is this what we really want dictating the services we use day to day?
Google tried to add support for it in their product
Is like saying that google tried to add support for HTTP to their products. Google Talk was initially a XMPP chat server hosted at talk.google.com, source here.
Anyone that used Google Talk (me included) used XMPP, if they knew it or not.
Besides this, it’s only a story of how an eager corporation adopting a protocol and selling how they support that protocol, only to abandon it because corporate interests got in the way (as they always do). It doesn’t have to be malicious to be effective in fragmenting a community, because the immense power those corporations wield to steer users in a direction they want once they abandon the product exists.
That being said, if Google Talk wasn’t popular why did they try to axe the product based on XMPP and replace it with something proprietary (aka Hangouts)? If chat wasn’t popular among their users, this wouldn’t of been needed. This could of been for internal reasons, it could of been to fragment the user base knowing they had the most users and would force convergence, we really can’t be sure. The only thing we can be sure of is we shouldn’t trust corporations to have the best interest of their users, they only have the best interest of their shareholders in the end.
These are just my opinions on the matter at hand.
TLDR; it’s not all about growing as massive as possible and letting everyone talk to everyone. It’s about communities being able to make choices for their user base and the freedom to choose who to federate with. It’s also about users having a choice of which instance they use to interact with the fediverse, and with whom. Having Meta involved limits these choices in not so obvious ways.
Doesn’t the fediverse have an inherent protection and/or immunity from corporate take-over?
Yes, but that does not mean it is invulnerable. Take the World Wide Web as an example, over the past couple of decades the decentralized web has become increasingly centralized. Projects such as Lemmy and Mastodon are a shot back at this trend, to try and break the web up as it was. Each instance gets to decide if letting large corporations federate with them is the best choice or not. It seems that a lot do not want this, and this is exactly the kind of protection from corporate take over that is inherent. The more large central servers are allowed to take a central role, the more power they will gain to snuff small communities and instances. They will do this by fragmenting users bases and communities over time, or any other dirty tricks they can come up with.
Also, having billions of dollars at your disposal is known to increase your influence overall. They can outspend anyone to sell most people on how Threads is interconnected and fediverse friendly, if you let them sell that lie they will win in time. They’ll do this, pull the rug and say how other independent instances aren’t corporating. They will shut off access to these communities in one way or another and begin the process of centralization. It has happened before, and will happen again.
Aren’t we protected?
If you choose to not use Threads, you are not giving your information directly to Meta. But, that does not mean you are safe. Meta is a corporation, and will try to pull whatever tricks they can to take over as the dominate player. They are going head to head with Twitter, what makes you think instances a fraction of Twitters size are safe?
Also, saying we are isolated by our individual instances is a bit humorous as they are federated. If one instance pushes most of the content is that really isolated? What about upvotes, engagement and any other activity that is pushed to other servers via the ActivityPub protocol? These will all be taken in by Meta, which means you are feeding them activity. Sure it’s safer, but they are still getting more data by engaging in the ActivityPub protocol than they get via scrapping pages. Also, they don’t have to play fair with the ActivityPub protocol, there are a lot of dirty tricks that could be used to hamper content on other instances than their own.
Is there anything currently stopping Meta from scraping the Fediverse for our content?
No, and the fediverse should not care. The goal of the fediverse at the moment is to stay independent and have a user base that is not reliant on a single entity and to stay away from the influence of corporate interests. If you operate in a public space, someones always going to be able to see it. It’s all about who owns that public space.
Won’t we grow & educate?
Who is we? Users that value their freedom will stay in the independent fediverse instances. Those who are looking for a twitter alternative will probably go to Threads. Those who don’t care will probably stay on Twitter. Any of these users might have multiple accounts on some or all of these services. Trying to group this together as “we” is a bit disingenuous.
As for growth, it’s not safe to assume that independent instances will grow because of the federation of users from Threads. Users that are on Threads are likely to stay on Threads, users that join instances are likely to stay there. Look to linux users to see why you aren’t going to convert many over the virtues of freedom and decentralization, you’ll just become another “fanboy”.
Aren’t we worried we’re forcing an ultimatum while the Fediverse is still in its infancy?
What is the ultimatum? This is a pretty loaded question, since some of the fediverse is already fractured. The fact you can spin up your own instance, invite whoever you want and keep the interests of your community out of the hands of corporations is the goal. Freedom to host your own community. Anything else is just having a capitalist mindset on growth, the line doesn’t always have to go up. Getting the most users isn’t the end game, it’s having a community that you belong to and feel a part of.
What’s the harm in pulling the ripcord if we try it, and it’s truly not a good fit?
Each instance chooses what is best for their community. Being a part of the mainstream content feed isn’t the goal of most of these decentralized communities.
“What about an influx of low-quality content?”
Why do instances need to let users block Meta when they know their users want Meta blocked? What’s stopping users from going to an instance that doesn’t block Meta if their instance disagrees with their opinion? It’s all about doing what instances communities want, or users can migrate if they feel their needs aren’t being met.
“What if Meta doesn’t moderate well?”
Meta will probably be able to moderate for their advertisers better than most instance operators will be able to. But again, it’s not about moderation and sanitizing content for advertiser revenue, it’s about having a space that is for the community by the community. It doesn’t need to be a single homogeneous community so ads can sell. Some of us want that outside of a corporations control, others don’t or don’t care, all are valid. Thankfully, everyone has a choice instead of being forced to do one or the other.
It’s still a fingerprint, the most vague information correlated with other data points can make a useful fingerprint. This is how a lot of the companies can track you even if you aren’t logged in, you using any service creates a pattern that with enough aggregate data can be used to approximate who you are.
Each instance serves the content from that instance, so from my understanding the only thing other instances can see are subscribed communities to be able to federate posts. Upvotes/Downvotes can possibly be tracked per user as they are federated on a per-vote basis currently, though this is just something I read and don’t have sources at the present.
I’ve never ran this program, but skimmed the documentation. You should be able to use the
SHIORI_DIR(or a custom database table following those instructions) along with the-pargument for launching the web interface. A simple bash script that should work:export SHIORI_DIR=/path/to/shiori-data-dir shiori serve -p 8081To run multiple versions, I’d suggest setting up each instance as a service on your machine in case of reboots and/or crashes.
Now for serving them, you have two options. The first is just let the users connect to the port directly, but this is generally not done for outward facing services (not that you can’t). The second is to setup a reverse proxy and route the traffic through subdomains or subpaths. Nginx is my go-to solution for this. I’ve also heard good things about Caddy. You’ll most likely have to use subdomains for this, as lots of apps assume they are the root path without some tinkering.
Edit: Corrected incorrect cli arguments and a typo.