Sublime
An inspiration engine for ideas
After 1 hr research,
Here are the best open source 'General agent' projects:
- Suna: https://t.co/BRsQToXL9P
- Deer-flow from Bytedance: https://t.co/4zwuRKaNFZ
- Google-gemini-search: https://t.co/iFIMgBxfeg
- Langchanin open deep search:... See more
This whole idea of having LLMs or an "agent" analyze, edit/refine, and expand your context makes a lot of sense.
Top-k/semantic search is a "weak retrieval" mechanism and inevitably leads to results that don't capture all the context needed to answer a question. Providing the tools to an agent to dynamically look up... See more
To understand the self, it's necessary to deconstruct our experience of subjective identity, and our experience of what it means for an object to be identical with itself. Identity is a representation, not a reality. Some represented objects have identity, others not.
Joscha Bachx.com
Getting started with thinking models + tools with a notebook and video:
I show off using the latest thinking models including Claude 4.0 and openAI 04-mini with tools from @tavilyai for web search and @ContextualAI for RAG.
To tie it all together, I use @AgnoAgi for a framework.
You can run it all for free... See more
„Zvykli jsme si na to, že drtíme nízkopříjmové a celý systém jede na nich, z minimální mzdy se jim vezme cca 120 tisíc ročně, oni se v systému nevyznají, tak si nestěžují a pomáháme těm bohatým. Je to stejné uplácení voličů jako když se přidává důchodcům", říká @dan_prokop https://t.co/DPEg1NoNKk
Google I/O. Some thoughts: the model seems to be multimodal in, but not multimodal out. Imagen-3 and music gen models are still detached from Gemini as standalone components. Merging all modality I/O natively is the inevitable future:
- enables tasks like "use a more robotic voice", "speak 2x faster", "edit this image... See more
Jim Fanx.comAI future
Josh Rayray • 1 card