

That’s true. I got to the op article through this one: https://www.der-postillon.com/2025/06/sonntagsfrage-tempo-30-berlin.html
Would this be a good fit? (I don’t speak German)
account feddit di @[email protected]
That’s true. I got to the op article through this one: https://www.der-postillon.com/2025/06/sonntagsfrage-tempo-30-berlin.html
Would this be a good fit? (I don’t speak German)
What are your interests?
In the meantime you could follow a couple of communities is where people share interesting videos:
(thanks @[email protected] for the correction)
Very cool!
I’m wondering how long until Bluesky locks down its users
Even if LLM “neurons” and their interconnections are modeled to the biological ones, LLMs aren’t modeled on human brain, where a lot is not understood.
The first thing is that how the neurons are organized is completely different. Think about the cortex and the transformer.
Second is the learning process. Nowhere close.
The fact explained in the article about how we do math, through logical steps while LLMs use resemblance is a small but meaningful example. And it also shows that you can see how LLMs work, it’s just very difficult
Yes, that’s it. I added the link in the OP,
Very arrogant answer. Good that you have intuition, but the article is serious, especially given how LLMs are used today. The link to it is in the OP now, but I guess you already know everything…
Thank you. I found the article, linkin the OP
Oh wow thank you! That’s it!
I didn’t even remember now good this article was and how many experiments it collected
I’m aware of this and agree but:
I see that asking how an LLM got to their answers as a “proof” of sound reasoning has become common
this new trend of “reasoning” models, where an internal conversation is shown in all its steps, seems to be based on this assumption of trustable train of thoughts. And given the simple experiment I mentioned, it is extremely dangerous and misleading
take a look at this video: https://youtube.com/watch?v=Xx4Tpsk_fnM : everything is based on observing and directing this internal reasoning, and these guys are computer scientists. How can they trust this?
So having a good written article at hand is a good idea imho
What are your interests?
Anyway you could try following this community: https://lemmy.world/c/peertube
Woah, I think you catchet the biggest problem I have with it immediatly. I’m thinjking of opening an issue maybe?
We both know that is not how this should work. If everybody joins the same instance
But this doesn’t happen. Mastodon.social is far from 90%, and the former CEO made the choice of letting control go to an entity that is no-profit, I think it’s the right choice.
I’m not sure, but i guess that it is replicating Twitter behaviour. I got used to that quickly.
But if you want a better experience you can easily find other front ends. I strongly suggest Phanpy that makes it really clear when a toot is part of a conversation/thread. It also does a lot of nice things like grouping boosts in horizontal carousels, collapsing long threads to avoid occpying the too many screens, and others, but it is still simple
I think your steps from 2 to 8 are summed up in
Create an account. Just like Twitter (or any other social media).
Meaning that those are not any different than what you need to do in Twitter or facebook. Actually the consents you need and the app permissions are much simpler in Mastodon case
By the way I think the communication can be much simpler:
Download the Mastodon app from the Apple app store or Play store.
Create an account (just like Twitter or any other social media)
Follow people, and get posting.
I have no fucking idea.
A group of Italian women who attended the event abandoned their short effort to join the platform half-way into the presentation, confused about where to go and which app to use.
Ouch!
Futhermore, comments aren’t grouped in the post. They are separated from their parent element. And you need to crawl back the comment to find its original post.
if you use a client like Phanpy things become much better from that point of view. Strongly suggested
I think that the lightest you can get is SNAC: written in portable C, it requires only an ngix server, and no database.
It’s ideal for single used instances or automated projects
Here some background and how tos: https://encrypted.tesio.it/2024/12/18/how-to-run-your-own-social-network.html
And here the description of fedimeteo, that provides forecasts for thousands of cities on the fediverse using a slim low cost vps’ https://it-notes.dragas.net/2025/02/26/fedimeteo-how-a-tiny-freebsd-vps-became-a-global-weather-service-for-thousands/
I have some favorites
https://makertube.net/accounts/michellemay about stichong, sewing etc
https://subscribeto.me/accounts/gbryant technology and videogames
https://video.canadiancivil.com/accounts/shifter active mobility and biking
Done.
Reminder for myself: read rules of community before posting, next time 🙂