One pattern I find useful for working with LLMs is a nice long ramble session. Sometimes the LLM needs more bits to understand what you're trying to achieve, but you're too lazy to type them. In these cases I like to lean back, switch to /voice and just ramble for like 10 minutes, total mess, anything goes, full stream of consciousness. Sometimes I declare it up top, something like "switching to speech recognition sorry for any typos...". Sometimes I turn it into a small interview of a few turns. But I find that the LLMs are somehow very good at reconstructing long incoherent rambles and often their echo of your own tangle of thoughts comes out quite a bit cleaner than what you started with. The result is that you improve the mind meld and have to correct things less from that point on.
To summarize: HuggingFace got autonomously compromised by a model from an American company. HF then tried to use American frontier model(s) to defend themselves, but were blocked by guardrails. HF then had to turn to open source Chinese models to defend themselves from another American company.
letting 5.6-Sol audit my entire agent stack top to bottom across 6 different machines. 4 days in now and the amount of optimizations it has identified is wild.
been fighting a bambulabs bug for 3 months, throwing every single agent at it with no progress
fable just hacked its way into the bambu firmware and fixed everything in a couple hrs
lightweight cli tool i made for post production workflows and managing media
offload cards, organize destination folders, generate proxies, and set up your project file all with one command
https://t.co/LY75Kh6Uj3
hermes agent made a custom websocket so we can work together in touchdesigner
it wired together most of this tree, then logged how i set up the composite nodes and turned it into a skill... crazy