As I rely more on self-hosted agents with OpenClaw and Hermes that I interact with messaging apps on my phone, I find myself wanting more UI affordances invented in first wave of chat bot exuberance: the WeChat bottom tab menus, Messenger suggested replies/carousels, etc.
Dan Grover
Verified account This account is verified. Learn more@DanGrover
5K Following 12K Followers 14K Posts
Latest posts 5
Currently printing parts for an Aloha Mini 2. Excited to experiment with it!
We got OpenWAM running on AlohaMini! 🤖View quoted post on X
Fine-tuned with just 90 demonstration episodes, our low-cost, open-source dual-arm robot achieved ~60% success rate on a real-world tangram puzzle task.
Full execution at 1× speed. No fast-forwarding.
#OpenWAM #EmbodiedAI #Robotics
One thing I've noticed for myself and friends is when in a dysfunctional/incompetent environment, one might know what the actually-right thing to do in a given situation is and self-censor by imagining a chorus of the most mediocre halfwit criticisms of that action that would inevitably come from your colleagues. 75% of the time, this is a trick: the lower the standards are, the more you are rewarded for still doing the right thing.
Some months ago, I felt purposelessness in face of all of the things that took me so long to learn that AI can do. This was epiphany that got me past it. Only way forward is to be as “out of distribution” and unapologetically unique in ways that might have seemed scary and risky in the old world.
your edge is whatever you know that the models don't knowView quoted post on X