Spyke

Replies

lemmy

Comment on

What's the term for Lemmy users?

All of the more fediverse-focused ones (fedditors, fedinauts, feddies, fedizens, etc) are definitely better. Lemmy is bigger than just Lemmy: we have users from Mastodon, Kbin and more seeing and replying to our posts. Let's choose a name that reflects that :)

Also, just to prove a point, if you're here from anywhere not Lemmy, say hello!

Comment on

How do we deal with similar communities on different Lemmy instances?

Reply in thread

I hear you man. I went from active contributor to mostly lurking on Reddit, and it wasn't even a conscious choice. Gradually, everything became very mechanistic. I knew what the top few comments would be before going to the comments. The churn became cyclic in nature.

After just a few days here, it was actually a little disconcerting how antagonistic and hostile people there are in the comments section. That's just how people communicate, on a hair-trigger from flamewar.

I recognize your username, I saw what you wrote about SQL scaling. Can you imagine recognizing a username in a major subreddit in the reddit of today?

The dichotomy between the big communities which people subscribe to from all over Lemmy and the small meta/announcement/server issue communities for each individual instance is gonna be interesting to see develop as the userbase increases. Kinda like the difference between seeing people from your street everyday, then many more less familiar people in the city center.

support

Comment on

five more new Beehaw communities!

Hi all, I know I'm not from Beehaw, but I do subscribe to a lot of your communities as I like your moderation.

Any chance of a World News community? The current News community is very US-centric, and it'd be nice to have a more international spot to put news.

Comment on

How has ur lemmy experience been so far?

Reply in thread

That's because they're different things in the fediverse.

Lemmy upvotes are mapped to ActivityPub upvotes

Kbin upvotes are mapped to ActivityPub boosts (basically a retweet in twitter language)

This also means that Kbin users have to be much more careful with what they "upvote" as that causes the post/comment to be shown to everyone that follows them in the wider fediverse. They are looking to change this if I recall correctly, and there's a lot of communication between the devs of Lemmy and Kbin on how to approach both this issue, and unifying their API system so apps can work for both.

support

Comment on

five more new Beehaw communities!

Reply in thread

Well, gave it a shot. We'll see what happens :)

EDIT: Seems to be performing about average, comparing it to other posts on the community. Ok, point retracted. Just need more people putting stuff in from outside the US. Maybe I'm just still stuck in the Reddit "News means US news" mindset and need to break out of it.

lemmy

Comment on

What's the term for Lemmy users?

Reply in thread

Do you mean at the technical level or UI level?

At the UI level, just have a look at https://kbin.social and you'll see a lot of very familiar posts. Also their posts show up here, just like any normal post.

Mastodon is a bit trickier. If you put hashtags in your post text, I think it shows up like a "toot" for users following it? Sorry, not a Mastodon user, don't really know how it works.

At a technical level, Lemmy is built on the ActivityPub protocol. It's how Lemmy servers talk to eachother, it's how Mastodon servers talk to eachother, and it's how a hell of a lot more services talk. Best analogy I've heard is ActivityPub is like the email protocol for social media platforms.

Comment on

Federation issues

Reply in thread

It still helps a lot because most of the load is coming from serving up the feed. Just federating the data to another server is much lighter than serving up everything via the instance's website.

news

Comment on

Don’t Worry, We’ll Kill Ourselves Before AI Can

So, there's a fundamental issue here. A lot of the systems that Amanda is talking about aren't actually AI.

Chat-GPT, contrary to the blogosphere, is not actually AI. It does not have the capability for thought. It doesn't have the capacity to understand truth or fiction as concepts, let alone tell them apart.

Chat-GPT and similar systems are probabilistic language models. Essentially, I start it off with sentences (a list of tokens, if you want to get technical). Then it responds by essentially looking at the training data it's been supplied with and picking out the sequence of tokens that most likely is the answer the user is expecting, given the input. Notice that bolded text? The user is expecting. Not anything else. These language models are trained to spit out what users expect, nothing more, nothing less. If a user doesn't like the response, they give a thumbs down and the model recalibrates, introducing more noise and randomness into the result.

These language models are actually really great at reducing manual labor at certain tasks (writing cover letters, delivering predictable essays, I've personally used Chat-GPT for Shadowrun world-building) but they need to have a knowledgeable person using them because they absolutely will not reliably say true things. They will say whatever their training data says is the most likely thing the user is asking for.

Comment on

How has ur lemmy experience been so far?

Reply in thread

I might try install an instance on my website and try to make a merge function

Awesome! I'm trying to get my feet wet with contributing as well. I don't know Rust or Psql very well (although I am an experienced MySQL/MariaDB admin) so it's gonna take a while for me to catch up enough to be useful. I'm trying though :P