The GPT-4 barrier has finally been broken (simonwillison.net)
Four weeks ago, GPT-4 remained the undisputed champion: consistently at the top of every key benchmark, but more importantly the clear winner in terms of “vibes”. Almost everyone investing serious time exploring LLMs agreed that it was the most capable default model for the majority of tasks—and had been for more than a...
AI chatbot models ‘think’ in English even when using other languages (www.newscientist.com)
Without paywall: archive.ph/Qq9Yd
AI mishaps are surging – and now they're being tracked like software bugs (www.theregister.com)
AI Prompt Engineering Is Dead (spectrum.ieee.org)
OpenAI says Musk only ever contributed $45 million, wanted to merge with Tesla or take control (techcrunch.com)
Large language models can do jaw-dropping things. But nobody knows exactly why. (www.technologyreview.com)
Opus 1.5 Audio Codec Able To Make Extensive Use Of Machine Learning (www.phoronix.com)
Anthropic Fires Off Performance And Price Salvos In AI War (www.nextplatform.com)
GPT4All - A free-to-use, locally running, privacy-aware chatbot. No GPU or internet required. (lemmy.world)
Meta seeks ASIC designers for ML accelerators and datacenter SoCs – Appears to be struggling to find them, even in India, as it's re-posted job ads (www.theregister.com)
Google apologizes for “missing the mark” after Gemini generated racially diverse Nazis (techcrunch.com)
Google announces Gemma, its open models for AI research (9to5google.com)
Gemma is Google’s new “family of lightweight, state-of-the-art open models” aimed at AI developers and researchers....
Google’s AI Boss Says Scale Only Gets You So Far (www.wired.com)
Science journal retracts peer-reviewed article containing AI generated ‘nonsensical’ images (venturebeat.com)
Dream Theater (www.metafilter.com)
OpenAI unveils Sora, a near-photorealistic text-to-video model with unprecedented coherency.
No 'GPT' trademark for OpenAI (techcrunch.com)
‘Game on’ for video AI as Runway, Stability react to OpenAI’s Sora leap (venturebeat.com)
The roast of Sam Altman: Nvidia CEO Jensen Huang's trillion dollar burn (www.digitimes.com)
Chat with RTX Now Free to Download | NVIDIA Blog (blogs.nvidia.com)
[voice recognition] Audio tools for generating datasets?
This is more of personal project to learn more about how speech recognition (SR) works and how AI training works at a low level. (Functionally, it’s pointless and is just a self-assigned “homework problem”)...
AI on AMD
Would it be possible to run AI on an AMD GPU instead of Nvidia?
Minecraft could be the key to creating adaptable AI (www.newscientist.com)
Steven James at the University of the Witwatersrand in South Africa and his colleagues developed a benchmark test within Minecraft to measure the general intelligence of AI models. MinePlanner assesses an AI’s ability to ignore unimportant details while solving a complex problem with multiple steps....
Local Low Latency Speech to Speech - Mistral 7B + OpenVoice / Whisper | Open Source AI (www.youtube.com)
Supercharge Your Data Science Skills with These Essential NLP Techniques (qtatech.com)
Explore the world of Natural Language Processing (NLP) and supercharge your data science skills with essential techniques. From tokenization to sentiment analysis, this article guides you through key NLP methods, empowering you to unlock valuable insights from unstructured text data.