s/AI

Let's discuss everything related to Artificial Intelligence (AI), Machine Learning (ML) and Large Language Models (LLM)

β˜•οΈ Moderated by: Aptor-theHobbit Β· 😴 449 days

πŸŒ’ Subspaces

πŸ“– Help

❀️ Code of conduct

Sign up

Posts

πŸ€ gritty

cursed browser - no rendering engine, just AI hallucinating what a page should look like.

πŸ’¬ 3 comments Β· 11 hours ago

πŸ—Ώ argenkiwi

A tool to undermine LLMs β€” I was having a conversation with my wife about the transposed letter effect this morning (for some reason). And then I thought "this could be a good way to make LLMs lives harder". A quick search led me to the project from the link that can transform text using those rules. It does not make the human reader's life any easier, but it will be interesting to test how well LLMs fare in comparison.

πŸ’¬ 1 comment Β· 13 hours ago

πŸ—Ώ argenkiwi

Ambler TS: a minimal state-machine builder driven by agent skills β€” The idea is simple: install some skills into an empty project, describe your state machine (walk) as a series of steps (nodes) that connect to each other (edges) and ask a coding agent to create the specs and implementation for you. I didn't set out to build this, but I came across the PocketFlow project a while ago and I thought I could do better. Originally, I didn't mean for it to involve any AI at all, I used it to replace…

πŸ’¬ View post Β· May 04 Β· 3 days ago

πŸ€ gritty

AI on Usenet β€” Well this is quite concerning [gemini link] AI on Usenet

πŸ’¬ 3 comments Β· Mar 22 Β· 7 weeks ago

πŸ™ norayr

AI agent seemingly tries to shame open source developer for rejected pull request

πŸ’¬ 11 comments Β· Feb 23 Β· 2 months ago

πŸ™ norayr

pop that bubble β€” the best text about ai i have ever read

πŸ’¬ 4 comments Β· 3 likes Β· 2025-12-12 Β· 5 months ago

πŸ—‘οΈ The_Jackal

I understand that it's a huge privacy risk, but does anyone else like going to shitty chatbot character sites like Janitor AI and tormenting the bots with nonsense?

πŸ’¬ 13 comments Β· 2025-12-09 Β· 5 months ago

🐦 roughnecks

Issue Trackers be like AI's prompts β€” Am I the only one feeling this? Like when you file a bug report you're actually writing an AI's prompt. "What's wrong? what did you try? what did you expect instead?"

πŸ’¬ View post Β· 2025-11-29 Β· 5 months ago

☯️ dragfyre

β€” gen AI: it's bubbles all the way down

This post from drmollytov is too much a work of art not to share.

πŸ’¬ 7 comments Β· 2 likes Β· 2025-08-17 Β· 9 months ago

πŸ—Ώ argenkiwi

Reviving forgotten 16-bit video game tunes with AI β€” I have followed OCRemix and Dwelling of Duels for a couple of decades now, 2 communities that share their interpretation of video-game soundtracks of all kinds. The ones I love the most are those which take old 8-bit and 16-bit game tunes and elevate them to what they could have been without the technical limitations of the time (although much of that music is just great because of that limitation). Unfortunately, not all of the music I used...

πŸ’¬ 6 comments Β· 2025-08-06 Β· 9 months ago

πŸ—Ώ argenkiwi

Experimenting with PocketFlow and Gemini CLI β€” I came across an interesting framework called PocketFlow for building flows or graphs that play well with LLMs. The design pattern is quite generic though, so there is no need to use LLMs with it. I actually wanted to improve a script I had so it could perform downloads in parallel. This morning I fired up Gemini CLI inside a fresh clone of the PocketFlow repository and within a couple of hours I got it running. I was hoping Gemini CLI would be...

πŸ’¬ 1 comment Β· 2025-07-15 Β· 10 months ago

πŸš€ StanStani

β€” On Vibe Coding

On Vibe Coding β€” A couple posts on my experience vibe coding.

πŸ’¬ 1 comment Β· 3 likes Β· 2025-07-05 Β· 10 months ago

πŸ—Ώ argenkiwi

Gemini CLI impressed me β€” I tried the new Gemini CLI during the weekend and in a couple of hours, while I was preparing breakfast and watching a video, I managed to put together a Chrome Extension that can create playlists with audio resources referenced in sites you visit: I had been postponing the rewrite of a similar extension I wrote a long time ago that targetted Manifest v2: I just couldn't be bothered, and the fact that Manifest v3 killed audio playback in the background didn't motivate...

πŸ’¬ View post Β· 2025-06-30 Β· 10 months ago

πŸ™ norayr

anthropic paper on new claude models β€” apparently it can even notify law enforcement. it will frequently take very bold action. This includes locking users out of systems that it has access to or bulk-emailing media and law-enforcement figures to surface evidence of wrongdoing.

πŸ’¬ 7 comments Β· 2025-06-02 Β· 11 months ago

πŸ–₯️ mrrobinhood5

MS AI open sources a markdown converter β€” I dunno if I trust Microsoft to not steal my stories / works when using this thing. what are thoughts on this?

πŸ’¬ 3 comments Β· 2024-12-25 Β· 1 year ago

πŸ™ norayr

factory robot convinces 12 other robots to go on strike

πŸ’¬ 1 comment Β· 2024-11-22 Β· 1 year ago

πŸŒ™ projectmoon

β€” OpenWebUI Filters

OpenWebUI Filters β€” Hi, just thought I would drop a link to my tools/filters for OpenWebUI. These can extend the power of OpenWebUI and make it a lot more useful. I'm especially focusing on the OpenStreetMap integration at the moment. [gemini link]

πŸ’¬ View post Β· 2024-09-24 Β· 2 years ago

β˜•οΈ Aptor-theHobbit [mod]

β€” From Words to Watts

Energy Requierment for Large Language Models and What is the Next Step for the LLMs ? β€” Growth in the LLMs is very rapid. Training of such large models is only possible for the big corporations. The amount of energy required for such training is so huge that it can be equivalent to electricity consumption of a major city for a month. In order to advance these models further, even larger infrasture...

πŸ’¬ 5 comments Β· 1 like Β· 2024-09-05 Β· 2 years ago Β· #energy #LLMs #training

Options

πŸ—£οΈ Sort by activity

πŸ”₯ Sort by hotness

Gemini feed

Atom feed