I'm pretty sure Claude 4 got nerfed.
LLMS
-
Minimax Hailuo-02 AI Model Now Available on Replicate
By
–
Here you gohttps://t.co/z8xyIA5TwG https://t.co/WorYHOko5o
— fofr (@fofrAI) 8 juillet 2025Here you go https://
replicate.com/minimax/hailuo
-02
… -
LLMs as Platform Technology: Rejecting Update Frequency as Primary Metric
By
–
i reject the assumption but even granting the assumption, frequency of update is the completely wrong metric to optimize/decide on here. llms are platform tech and need to be on as low a level as possible or we will just have electron-but-llms all over again.
-
xAI Employee Teases Grok 4 Livestream Announcement for Wednesday
By
–
Un employé de xAI sur grok 4 : Traduction : " j’étais absent, j’étais occupé à écouler des GPU. Je rentre bientôt. Direct (livestream) Grok 4. Mercredi. 20h."
-
OS-level LLM integration instead of browser bundling
By
–
all of these are stupid tbh. bundling browsers with llms are stupid. distribute it with the os and let the browsers hook into them like a syscall.
-
Chrome 138 Ships Gemini Nano Locally to 3.7 Billion Users
By
–
so Chrome 138+ onwards ships Gemini Nano for every user*, putting a local LLM in 3.7 billion monthly active users of Chrome wrote a personal guide today for building with Gemini Nano – mostly for me, might help you if you're not suuper strong at javascript or parsing google's…
-
Chatbot AI Transparency: Users Demand Clear Disclosure Requirements
By
–
I am fine with being connected to an LLM-powered chatbot for support – sometimes they are really good I am not fine with not being told I am being connected to an AI chatbot for support (or a tiny "powered by BlongoMax" or whatever logo like I am supposed to know what they are)
-
Prompting AI Assistants for Learning and Innovation
By
–
One of the key problems with AI right now is that "helpful, friendly assistants" are not the personality you want when you are trying to learn, trying to innovate, or trying to get the best work out of a group. But a little bit of prompting goes a long way, see our tutor prompt.
-
GPT 4.5 Discontinuation Announced: Service Ending in One Week
By
–
reminder GPT 4.5 dies in 7 days. i turned it off today
-
GPT 4.5 Underperforms: Reasoning Models Validate New AI Paradigm
By
–
as the sun sets on GPT 4.5, just reflecting that it was really not that great for summarization tasks. today's @smol_ai example here i think the "failure"* of GPT 4.5 relative to o3/o4mini is actually fantastic validation of the bet on the reasoning paradigm. 10-100x larger