Martineski

joined 1 year ago
MODERATOR OF
[–] Martineski@lemmy.fmhy.ml 6 points 1 year ago (3 children)

Yeah, you're right. I assumed that he may have given the GPT some key points and made it tell all those things in a certain tone but I was def reading too much into this.

[–] Martineski@lemmy.fmhy.ml 2 points 1 year ago (5 children)

Isn't it GPT generated? Just guessing because it feels very weird to read lol.

[–] Martineski@lemmy.fmhy.ml 2 points 1 year ago* (last edited 1 year ago) (8 children)

Alright, my response is... Tags won't function like mastodon and such, each community will have their own tags to further categorise content of a general topic that sublemmy is about. (Edit: actually, I think it may be do BOTH of those things now that I read my conversation with the dev again)

This is what the dev said about stages of tag system (in my conversation with him on dc):

Stage 1: Tagging and Presets (Community and Instance, no User defined Tags allowed)

Stage 2 (not sure if I'll do that): Filtering, Cosmetics? (Border, Text Color)

And this is what he told me when I asked about the possibility of stage 3 of the tag system where you can subscirbe to only certain tags so only choosen content from a community shows up on your feed:

I think backend wise not worrying about that is fine for now, the separation of tags per community can be achieved by filtering tag AND community ID in the filter, which could just be hidden from the user or displayed, there's a separate discussion to be had about the filter features. Point is not separating these tags per community shouldn't limit their filterability because community id's already exist. Plus by not splitting them up in the backend general tags can be used across instances a lot more easily.

Subscribing to tags in a community could then work similar to that filter, again the issue would be how to make the UI bearable but backend wise should be no problem to implement.

So my idea is possible from what he told me. The problem is that we need to make the idea known and we need to find someone who will work on it because the dev isn't even sure if he will be the one that implements stage 2 when I'm talking about stage 3. Also, to implement my ideas for stage 3 we need foundation to be laid out which means that the stage 1 and stage 2 need to be completed first.

For the stage 3 to be implemented (if anyone decides to implement this AT ALL) we will need to wait for weeks at best and months at worst. Now this makes it hard for me to decide how to proceed with my communities because the path of the tag system is not yet clear and I need to decide NOW which communities should be separate and which should become more general.

Edit: I talked with the dev, it will take months to reach stage 3. And I realised just now that not all content on c/imaginaryaviation will be vehicles which means that I will need to a create a new sub called c/imaginarylocomotion to be able to fully capture both of the subs.

[–] Martineski@lemmy.fmhy.ml 2 points 1 year ago (10 children)

I need to mentally prepare my ADHD brain for this before reading it xD. I will go do something else before I get ready lol.

[–] Martineski@lemmy.fmhy.ml 2 points 1 year ago (16 children)

Can you tell me what exactly do you dislike? Becasue I will be discussing things with the person that will code it (if he responds).

[–] Martineski@lemmy.fmhy.ml 4 points 1 year ago

Looks like my attempt at being sarcastic failed miserably. Lesson learned xD.

[–] Martineski@lemmy.fmhy.ml 3 points 1 year ago (2 children)

What?! Where do you live?

[–] Martineski@lemmy.fmhy.ml 13 points 1 year ago

At least one person in the comments gave me a good feedback. But to decide what to do I will need to contact the person that works on tags on github and go into an in-depth discussion with him about how it will work when the first stage of the tags isn’t implemented yet…

[–] Martineski@lemmy.fmhy.ml 4 points 1 year ago (18 children)

Here's the github issue for this, I will be contacting the person that works on them later to discuss more advanced filtering and tag following after the basic tag functionality gets implemented: https://github.com/LemmyNet/lemmy/issues/317

[–] Martineski@lemmy.fmhy.ml 10 points 1 year ago* (last edited 1 year ago) (22 children)

I'm already trying to bulk up whatever I can into one subs because reddit imaginary subs have a sub for every damn niche thing lol:

https://www.reddit.com/r/ImaginaryBestOf/wiki/networksublist/

https://www.reddit.com/r/ImaginaryUnofficial/wiki/index/

I will move tanks sub and mecha sub into vehicles sub because they will be adding tags to the platform and unlike reddit you will be able to use multiple tags for one post so there won't be any need for dedicated subreddits because filtering should be easy enough, you're right. I think that I will leave aviation sub as it is though because it's deistinct enough to keep it as a separate sub. Thank for the feedback!

Edit: for example, not long ago I moved imaginary feels into more general sub imaginary emotions, so it's not like I'm not trying to fight those problems. It's just a bit confusing to make more general sublemmies because reddit had subreddit for each specific things and it's hard to shift to doing it in another way haha.

Edit2: yeah, I will need to rethink (again) how I will do things with with imaginary subs. I will stop posting temporarily on the ones that I'm not sure what will happen with them yet.

Edit3: Actually, maybe I will merge imaginary aviation and ships together but I'm not sure how well the platform will work if a community will have tens of tags.

Edit4: Another reason for why I didn't include everything under a more generalised sublemmies is that if a person subscribes to a community then it will show ALL the posts on it and not only the content that the person is interested it. I pitched an idea on github for advanced filtering options for tags, maybe I should also mention them that it would be good to have an option to subscribe to tags on communities so it doesn't show all the content from them. But that also sounds like something that will be needed to be done in the backend and will take a really long time to implement, ugh.

[–] Martineski@lemmy.fmhy.ml 4 points 1 year ago (4 children)

I know, the title is a bait. Look at what sub is this.

 

If you are lurker please help us grow the community by commenting and posting interesting on topic quality info/discussions so we can attract more people and make this community more interesting to spend time on. ✌️

Previous milestone: https://lemmy.fmhy.ml/post/389193

Next goals: 400, 550, 700, 850, 1000

 

If you are lurker please help us grow the community by commenting and posting interesting on topic quality info/discussions so we can attract more people and make this community more interesting to spend time on. ✌️

Previous milestone: https://lemmy.fmhy.ml/post/322564

Next goal: 550 subs

 
 
 
 
 
 
 

Copied from reddit:

Not a lot of new information, but more discussion on Gemini and what it is going to focus on. It sounds like there's an emphasis on combining Reinforcement Learning with LLMs, but also a few other advancements they are keeping close to their chest. Relevant excerpts:

But Hassabis says his team will combine that technology with techniques used in AlphaGo, aiming to give the system new capabilities such as planning or the ability to solve problems. “At a high level you can think of Gemini as combining some of the strengths of AlphaGo-type systems with the amazing language capabilities of the large models,” Hassabis says. “We also have some new innovations that are going to be pretty interesting.” Gemini was first teased at Google's developer conference last month, when the company announced a raft of new AI projects.

...

Gemini is still in development, a process that will take a number of months, Hassabis says. It could cost tens or hundreds of millions of dollars. Sam Altman, OpenAI CEO, said in April that creating GPT-4 cost more than $100 million.

I do wonder what they will introduce. Maybe a memory system like what was seen in Retro, or maybe even some sort of continual learning paradigm that makes it easier to update? Time will tell

 

In Toronto, a candidate in this week’s mayoral election who vows to clear homeless encampments released a set of campaign promises illustrated by artificial intelligence, including fake dystopian images of people camped out on a downtown street and a fabricated image of tents set up in a park. In New Zealand, a political party posted a realistic-looking rendering on Instagram of fake robbers rampaging through a jewelry shop. In Chicago, the runner-up in the mayoral vote in April complained that a Twitter account masquerading as a news outlet had used A.I. to clone his voice in a way that suggested he condoned police brutality. What began a few months ago as a slow drip of fund-raising emails and promotional images composed by A.I. for political campaigns has turned into a steady stream of campaign materials created by the technology, rewriting the political playbook for democratic elections around the world.

 

If you are lurker please help us grow the community by commenting and posting interesting on topic quality info/discussions so we can attract more people and make this community more interesting to spend time on. ✌️

Previous milestone: https://lemmy.fmhy.ml/post/347505

view more: ‹ prev next ›