bogtap82

I mean, "AI" as it presents itself in the zeitgeist is in its infancy. It's like making conclusions about the internet in the late 90s. Whatever it becomes likely hardly resembles what it is now. For better or for worse.

In some cases I would agree with this, but image model releases including this one are beginning to incorporate and market the thinking step. It is not a reach at this point to expect the model to take liberties in order to deliver a faithful and accurate representation of your request. A model could still be accurate while navigating your lack of specificity.

It was renamed many times. It was also called "clawdis" at one point, and prior to that "warelay," when it was simply a Whatsapp gateway for Claude Code. It was already gaining some momentum at that point but wouldn't reflect as search results for "Clawdbot," and especially wouldn't be visible on Google Trends when most of the conversation was on X/Github.

I cloned the clawdbot repo back when it was named warelay or clawdis, can't remember, but it was much less dense then. Mainly cloned it for the in-the-box Whatsapp implementation. Since then I've built it into a pretty awesome agent for my home and family, who all have their own privileged access which allows it access to different skills and a mixture of shared and personal information. I have no interest in reconciling the Frankenstein I've built with newer mainline features, but the custom nature of my build is one of the things I find so fun and helpful about it. It's become so much more "mine" by just asking it to build out xyz feature for itself, and now it can do a bunch of weird things that revolve around its persistent access to the information I provide it and my ability to interface with it through a regular messaging app.

There is one single app I've been able to find that offers Parakeet-v3 for free locally and it's called Spokenly. They have paid cloud models available as well, but the local Parakeet-v3 implementation is totally free and is the best STT has to offer these days regardless. Super fast and accurate. I consider single-user STT basically a solved problem at this point.

At this point I really don't want to even imagine the headache of implementing their codebase updates into all the custom scaffolding that my own fork relies on now. I think my plan going forward is to cherry pick features that sound interesting and re-implement on my own using my agent that has proper documentation on my personal configuration. Will check out what's new though.

I've been running a variation of this for the past 3 weeks. I swapped out the default pi agent back to Claude Code because I didn't like the smaller feature set. Bought a phone line and communicate with my agent via iMessage on a clamshelled mac. A Tailscale network connect the head agent to all the computers on my network including my laptop, a few raspberry pi's, steam deck, and all the IoT devices in my house. As I discover new uses, I ask it to make skills and it is remarkable what it's been able to handle all through the single chat interface because it has 24/7 access to all my computers' file systems and my home network. It's been really fun to see how far I can take it, and the skills framework built into CC/Codex now make it feel infinitely extensible.

Thanks, I'll let them know.