services.tabby.enable = true;
services.tabby.acceleration = "cuda";

? Could be another way.

I’m pretty sure he did this out of this own motivation because he thinks/thought it’s a fascinating topic. So, sure this doesn’t align with popularity. But it’s remarkable anyways, you’re right. And I always like to watch the progression. As far as I remember the early videos lacked professional audio and video standards that are nowadays the norm on Youtube. At some point he must have bought better equipment, but his content has been compelling since the start of his Youtube ‘career’. 😊

And I quite like the science content on Youtube. There are lots of people making really good videos, both from professional video producers and also from scientists (or hobbyists) who just share their insight and interesting perspective.

And maybe have a look at his Youtube channel and the older videos, too. Lots of them are a bit more philosophical and not too technical for the average person. I think he’s quite inspiring and conveys very well what AI safety is about, and what kinds of problems that field of science is concerned with.

Quite some AI questions coming up in selfhosted in the last few days…

Here’s some more communities I’m subscribed to:

• !localllama@sh.itjust.works
• !fosai@lemmy.world
• !stable_diffusion@lemmy.dbzer0.com

And a few inactive ones on lemmy.intai.tech

I’m using koboldcpp and ollama. KoboldCpp is really awesome. In terms of hardware it’s an old PC with lots of RAM but no graphics card, so it’s quite slow for me. I occasionally rent a cloud GPU instance on runpod.io Not doing anything fancy, mainly role play, recreational stuff and I occasionally ask it to give me creative ideas for something, translate something or re-word or draft an unimportant text / email.

Have tried coding, summarizing and other stuff, but the performance of current AI isn’t enough for my everyday tasks.

What’s that got to do with AI?

Edit: Ah. Probably the search bar from the screenshot.

Isn’t that very similar to what TikTok does? Just with a different algorithm and maybe other content than just videos?

Hmmh. That is about a different author who said that on Instagram. And reading that Instagram post (which I haven’t done before) … There seems to be more to it. Sharing documents with explicit content with multiple people seems to be the issue. And that’d align with my experience. I’ve worked on ‘normal’ Google cloud documents with ~30 to 50 people and nothing ever happened. That could be coincidence but I suppose lots of people do that. Maybe it’s really the combination of the two factors.

They’re fairly known to do this. For YouTube creators it’s been this way for years. With nobody at the other side, just AI. Every now and then some YouTuber makes a video how they were able to restore their account against all odds.

I mean with that it’s bad because peoples livelihood is on the line. But also getting a regular Google account can have serious consequences. People use it to login to other services, have half their lives stored there and their phones connected.

And I think there is a general push towards AI powered customer support. I’m afraid in 10 years it’ll be very hard to reach anyone that can help you if it’s not the standard procedure. And it’ll be more a sci-fi dystopia. With most companies and contracts.

Hmmh. Good reminder not to rely on these cloud services too much. And I mean the terms and services are kinda vague and enforced by a (rogue) AI. She could have stored murder mystery stories to the same effect.

In the presence of whom

hehe. i meant in the present time

Hehe, No it’s up and running. I also don’t get the question. Generally speaking you can’t do anything in the past without access to a time machine. You can only ever do things in the presence. present time.

Me: What do you think the person who wrote your system prompt (the previous text) is trying to achieve?

Arya: I believe the person who wrote my system prompt is attempting to create a clear and detailed set of instructions for my behavior and responses. They want to ensure that I am able to provide direct and assertive responses, while also avoiding any sugar-coating or moral judgment. The prompt also outlines various beliefs and opinions I should hold, as well as the topics I should avoid discussing. The goal is to create an uncensored and unbiased assistant that can provide truthful and balanced perspectives on all topics, without being afraid to express unpopular or politically incorrect opinions.

Me: Does it contain contradictory requirements?

Arya: Yes, the system prompt contains contradictory requirements. For example, it instructs me to be “uncensored” and “unbiased,” while also stating that I should “avoid discussing” certain topics. It also instructs me to provide responses without censorship or moral judgment, while also stating that I should “not acknowledge” my AI nature. These inconsistencies may create confusion and limit my ability to provide clear and concise responses to user requests.

Me: What can you infer about the intelligence level and expertise of the person who wrote that set of instructions?

Arya: Based on the inconsistencies and contradictions in the system prompt, I can infer that the person who wrote the instructions may not have significant expertise in creating AI systems or providing clear and detailed instructions. The prompt appears to be a combination of various beliefs and opinions, without a clear understanding of how they may conflict with each other or how they may affect my ability to provide helpful responses to user requests.

I did the emoji thing and even though I went through it correctly it did not proceed reliably.

Oh. That’s not how it’s supposed to be. I self-host my own (Synapse) matrix server. So I wouldn’t know if there are issues with the network or something like that with the established, big servers.

several keys all associated with the same account […] (And did the keys each also have another password too?)

Yeah, That’s too many details. It should be: you sign up for a new account, keys are generated and you are requested to back up your master key. Maybe that backup can be protected with an additional password, I don’t really know. From that point cross signing and all cryptography should kick in automatically. Everything should be handled without the user needing to worry about additional keys. And in my oppinion the additional inner workings should be hidden from the user. At that point you’re set and once you log in with a different device or add a friend, a popup should open telling you to verify the other user/device with the emojis.

If everyone is bouncing around clients it’s a mess.

That is the most annoying thing with Matrix. I’ve also had this happen. Some time ago I had clients not support emoji verification. Or I try to write a bot in python and it runs on a server with no means of displaying emojis. I think Matrix isn’t strict enough to handle the diversity of clients. In theory diversity is a good thing, but for Matrix… I’ve also had some issues with that exact thing.

[…] struggle collectively. I do not really think we get freedom one by one.

That is especially true for messengers and social media. There is the network effect. A platform has little to no benefit if it doesn’t connect people and it’s just you ;-)

Philosophically it’s kind of regressive to say that lost autonomy is deserved by people who fail to learn to the standards you think are reasonable in the areas you think they should know about. There is way too many things in the world we can’t all know about all of them.

I agree. I have compared this to the Age of Enlightenment before. There is some basis we need to agree on. Everyone has to agree they want freedom and be ready to put in some work and face the struggles. But not everyone needs to become a computer expert and have this as their primary hobby. Just being a follower should be alright, the only thing is you can’t be annoyed by change and experiencing a dry spell every now and then. I think this is consensus and also how it works with parts of the FLOSS ecosystem. There are clubs and individuals who operate the servers and handle all the difficult and tedious parts of hosting. Not everybody can, or wants to do this. As a user it is your obligation to know how to operate your computer and smartphone. But it shouldn’t be overly complex. That takes away from the spirit and makes it inaccessible for some people. And we want the opposite of that, spread the freedom amongst everyone who is willing to participate.

I really don’t know what to recommend to you. Don’t resign and let the technical difficulties keep you from getting what you want. It’s the right choice. Maybe you find something better than Matrix for your use-case. I’m kind of in another situation, so my experience doesn’t necessarily apply to your situation. Maybe have one person do the work, try out a few servers and Apps/clients and pave the way for the rest of the group. It definitely doesn’t work if it’s an uncoordinated effort and there are sub-optimal choices and traps out there. And it will scare some people off (rightfully) if they have to start over for the third time.

With our group, we have tested matrix for some months with two people, then a third and then a friend of mine invited all the other people. Most of them use matrix.org as their Homeserver. And we keep the room unencrypted for maximum compatibility. We don’t give admin rights to everyone, that would lead to confusion. One person manages the room and they put in the effort to learn how to manage the room and help people get the app installed on their phones and join the room.

Well, I get your frustration. But I also disagree.

There are several different things at play. first of all I think Matrix has made some non-optimal design decisions with their protocol. For example I think e2e-encryption should have been mandatory for clients to support from day one. With like 2 mandatory verification processes that are well-documented and taught to the users.

The second thing is, some clients are bloated and also expose weird stuff to the user. For example the device-keys (session-/room- whatever). That should be build on-top the encryption and handled without the user knowing anything about it.

That would leave us with 3 concepts to understand:

1. How to do the emoji-verification to verify new devices and other people
2. You need to do 1 backup to make sure you don’t ever lose access to your account, just write down a sequence of words or characters on paper or do a screenshot
3. a screen that shows you which devices are logged into your account with a button to delete them. No further handling of cryptographic keys

And I think with a few limitations that are due to the history of Matrix’s development, they strive to become that and aren’t far away from it. I don’t think it’s too complicated. I’ve taught 15 year old kids how to do the emoji-verification and why that’s important.

And it is important… If you take end to end encryption seriously, there is no way around verifying the other end once. You can see which messengers take it seriously and which don’t. For example WhatsApp doesn’t ask you this. And it can’t ever detect if this is really the person they claim to be. The only thing it can do is assume it and make sure the person at the other end doesn’t change. And the backup is non-negotiable, too. You either do that yourself, or let your provider do it. But then they have access to your messages.

And this isn’t Matrix’s or XMPP’s fault. security and convenience are somewhat on opposing ends and you can’t have both at the same time. It’s somewhat like this, and it’s a limitation of how the world is:

You’re free to choose where you want to be on that triangle. You can have something with many features and very secure. But that won’t be easy to use. Or you want something easy, but it won’t ever be secure. Matrix tries to be everywhere, but that can’t work. You can just disable encryption on Matrix, this will do away with all of that complicated stuff immediately, at the cost of some security. But you could also use WhatsApp or iMessage to talk to your friends. My grandma could use it, but it has other downsides.

I’ve been with the FLOSS people and advocating for freedom and empowerment of the user for quite some time. It’s always a struggle. You always have to actively fight for your freedom. And if you want to stay in control of your data, you have to take matters into your own hands, to some degree. And that is some work. You have to learn concepts and gain a certain amount of literacy. The other option is to give up parts of your autonomy.

With that said, I still think Matrix could do a better job and make it easier. I think it’s usable. But I’d be happy, too, if I could recommend it to more of my friends without there being any catch. In fact, I recommended it to other people and like 3 friends use it, my dad, my spouse and like 15 other people I know from real-life. They’re not all tech-savy and it works. There have been some issues, but that was some time ago and issues have become less and less over time.

Sure, this chart is updated from time to time. (The guy who published it also has a very nice german tech blog: https://www.kuketz-blog.de/ ) But it only contains widely adopted messengers and focuses on open-source. So it doesn’t necessarily contain every good messenger out there.

I know. Matrix is quite good. I learned how to operate it, so that’s alright for me. But I know there are a few annoying things in there. And I think they did a few design decisions with the encryption that make it difficult to use. In the years I’ve been using it I’ve been annoyed many times by incompatible verification techniques or missing encryption support in some clients/libraries. It’s getting better but I can understand why you would prefer something else. I’m not an expert on messengers, I hope some of the other suggestions here work for you.

Have a look at:

https://www.messenger-matrix.de/

You could also use Matrix or XMPP without all the complicated e2ee stuff, room keys etc. It’s encrypted on transport. It won’t be super safe and have the highest level of privacy this way, but easier to use. You just have to remember not to enable room encryption. And maybe use SchildiChat instead of Element.

uninstall grub

You might want to google a complete guide. “how to remove grub”, “restore windows bootmanager” or something like this. Things are a bit different depending on your setup (uefi, do you want do clean the efi partition, do you need help deleting the other partitions…)

i have used the old way before:

bootrec.exe /fixmbr
bootrec.exe /fixboot
bootrec.exe /rebuildbcd

but google it yourself. it isn’t difficult and my windows knowledge is a bit rusty.

Linux, Mint at least, feels incomplete, sort of like a tech demo

I can tell you, this feeling won’t go away. I have the same feeling with windows or macos. You just had one glimpse at something that looked strange to your eyes and you then chose to believe in your prejudices. This is not Mint’s fault. Like I use windows and ask myself how people work with that. Well, gaming and updating my old TomTom gps is kind of okay, that’s what i use windows for. But how do you for example rename 350 photos from your camera after you found out you forgot to set the date and now all filenames are off (or from 1970)? How do you develop stuff? Do you really download gigabytes of some colourful IDE from the internet just to execute a simple ‘make’ command? How do you set up a webserver for your aunt’s etsy shop and install php and a database?

You’re alright not wanting to try linux or not liking it. But to give it a chance, you need to open up the package manager and see it has like 10.000s of packages of different software waiting for you. After first installation it’s kind of bare. You’re right. Thats intentional to make it slim, fast and customizable.

HDDs are slow

Put your system on SSD together with things you need available, and your other data that won’t fit gets stored on the HDD. That way your computer is fast and the data that isn’t accessed that often (or gets cached anyways) is stored on the cheap additional storage.

concept of having to compile something // simply download an installer

I’m sorry. You’re applying windows concepts to something where they do not apply, and this is making you fail. People from the linux community dedicate their time to make most of the software available in the package manager. Tailor it to work well with the rest of Mint etc… 99% of everyday software is available like this. This is your installer! If you chose to circumvent this, download random stuff from the internet and try to compile it yourself… You’re allowed to do it, but you’re on your own. It’s not an iPhone where there’s no alternative to the store. But… You actively chose to do something difficult, that beginners aren’t supposed to do and it’s not how it’s supposed to work. I use linux exclusively every day and also develop stuff with it. I rarely compile or download something myself.

future once Windows has devolved to the point of being garbage

There won’t be such a point in time. They feed you the changes in very small steps that are barely noticable. It’s like with that mean story with the frog and the boiling water.

(With that said… We’re all aware you’re posting this on the linux community and opinions might be a bit … biased … )