Introducing the ashen benchmark
A completely incorrect yet important way to keep score of new models on how good they are at the fun AI stuff. Over the past year or so, anytime a new model has come out, Ialways put it through a few specific, unusual tests: How good is it at 3D
Most AI apps do what they're built to do.
This one lets you change what it's built to do. Why DeepSeek Harness stands out: ✔ Open source, so anyone can see how it works ✔ Plugin manager shows what each add-on does and where it came from ✔ Creator mode builds the extras
Before the local, I created my APIs with #Deepseek.
It took time (creating folders, files, correcting...). Then, I tested the agentic #OpenClaw, #Hermes and it's exactly the kind of tool we need: an Orchestrator.
As I'm using the DeepSeek Harness, I'm starting to think, ``It would be a waste to use this as just a coding agent.''
What we are doing this time is an experiment in which we have DeepSeek Harness operate the browser, create a new chat in the ChatGPT Project, and automate the creation of articles.
You have reached the end of the archive
All of deepseek