hahaha fun photo. AndrejGPT-mini and a group of parallel subagents loading the context window with this 10,000 LOC codebase in gitingest, reasoning='low'
LLMS
-
GPT-5 Gets Warmer, Friendlier Tone in Latest Update
By
–
We’re making GPT-5 warmer and friendlier based on feedback that it felt too formal before. Changes are subtle, but ChatGPT should feel more approachable now. You'll notice small, genuine touches like “Good question” or “Great start,” not flattery. Internal tests show no rise in
-
GPT-OSS demonstrates impressive safety maximization features
By
–
Ohh nice to see how safemaxxed gpt-oss is!
-
vLLM Pull Request 22845 Merged Soon
By
–
That'd be cool! Here's the PR: https://
github.com/vllm-project/v
llm/pull/22845
… (but should be merged very soon) -
AI News Roundup: GPT4o Return, o3 Victory, Grok Free Access
By
–
A busy AI news week means an extra long AI news roundup post Did I miss anything? – @OpenAI restored #GPT4o as model option after #GPT5 backlash
– OpenAI’s o3 won Kaggle AI Chess Tournament
– @grok 4 is now free to use worldwide
– @NotebookLM Overviews went live for -

Claude Opus 4 gains ability to end conversations for welfare
By
–
As part of our exploratory work on potential model welfare, we recently gave Claude Opus 4 and 4.1 the ability to end a rare subset of conversations on http://
claude.ai. -

NinjaTech Builds Advanced AI Agent with Qwen3 235B
By
–
Today, using Qwen3 235B Instruct, NinjaTech has built one of the world’s most intelligent agents. In the challenging GAIA General AI Assistants eval, SuperNinja's custom scaffold scored in the same league as o3, Gemin 2.5 Pro and Sonnet 4.
-
NinjaTech Achieves One-Minute Frontier Research with Cerebras
By
–
In June, the NinjaTech team reached out to Cerebras with the ambitious goal of cutting deep research time down to 1min at frontier quality. We reached the speed, but the models at the time weren’t quite smart enough vs. o3.
-
SuperNinja: 5x Faster Deep Research AI Powered by Cerebras
By
–
Introducing SuperNinja – the world’s fastest deep research by @NinjaTechAI powered by Cerebras:
— Cerebras (@cerebras) 15 août 2025
‣ 1-2min for full answer instead of 10min
‣ 5x faster than o3, Gemini Pro etc.
‣ Same accuracy as measured by GAIA pic.twitter.com/rXDoamiksmIntroducing SuperNinja – the world’s fastest deep research by @NinjaTechAI powered by Cerebras:
‣ 1-2min for full answer instead of 10min
‣ 5x faster than o3, Gemini Pro etc.
‣ Same accuracy as measured by GAIA -

Character-Level Tokenization Success in DNA Language Models
By
–
watch me realize in realtime that @gdb is deadly serious about character-level tokenization and it's already working well for @pdhsu, @SKonermann, and @patrickc's @ArcInstitute 's 40B model that is the "GPT-1 of DNA"
— swyx 🐣 (@swyx) 15 août 2025
with biological language models (instead of natural language),… https://t.co/SMA81JYX0b pic.twitter.com/XXsMmDEZWewatch me realize in realtime that @gdb is deadly serious about character-level tokenization and it's already working well for @pdhsu
, @SKonermann
, and @patrickc
's @ArcInstitute 's 40B model that is the "GPT-1 of DNA" with biological language models (instead of natural language),