freeatlantis.com is one of the many independent Mastodon servers you can use to participate in the fediverse.

Administered by:

Server stats:

215
active users

#LLM

18 posts17 participants2 posts today

OK so the tech #industry now calls for the abuse of people's data for #AI training to be allowed under legitimate interest as legal basis under #GDPR. They really have no shame.

Quote: "The training of AI models should be able to rely on a legitimate interest basis, as provided for in the EU’s data protection regime GDPR, #IBM’s Chief Privacy Officer Christina Montgomery told Morning Tech."

@puppygirlhornypost2@transfem.social

The bad news - and there's a lot of bad news - but one problem that goes overlooked it how
#AI will further reinforce #epistemic #factionalism and segmentation. We're already seeing that with #socialmedia.

My social media feed is an intimate safe safe for progressive communities of anticapitalist resistance. Your social feed is misinformation and radicalization and propaganda.

Customized
#LLM prompts will crank that tendency past the breaking point.

Today in less time than it takes for #Ollama to provide me with a wrong answer I posted a question to the #Rust forum and got two, yes two, immediately helpful and correct answers, one including the code modifications I needed to make.

Problem solved by humans. 🙏

I know from experience that #LLMs would likely misunderstand the very carefully posed problem (with full code) and send me on a wild goose chase.

AI chatbots are like that one uncle you see at Thanksgiving. Over his life he became a true expert in something, say, ice sculptures, and can speak with authority on it. But he's convinced he should speak with authority on any subject at all — medicine, law, business, how rainy it will be next spring — stating with 100% certainty things he heard or made up.

Many things he says are right, many are wrong, and you can't judge which is which from his words alone.

#AI#GenAI#LLM

Where can I find The Internet™ without any #AI nonsense? It is getting harder and harder to find something as basic as coloring pages for my kids that don't have crazy-pants artifact tell tales. I want search results prejudiced against #LLM generated fluff, social media feeds without bots, a YouTube filtered for only human creativity, images of natural beauty—the natural real world with all it's warts, not the matrix.

Mistral.ai released an updated version of its Mistral-Large-Instruct model. The weights are available here:

huggingface.co/mistralai/Mistr

> Mistral-Large-Instruct-2411 is an advanced dense Large Language Model (LLM) of 123B parameters with state-of-the-art reasoning, knowledge and coding capabilities extending Mistral-Large-Instruct-2407 with better Long Context, Function Calling and System Prompt.

huggingface.comistralai/Mistral-Large-Instruct-2411 · Hugging FaceWe’re on a journey to advance and democratize artificial intelligence through open source and open science.