…figure out what you did to break things and get back on track.

Am back at being able to get up to speed with the Monocle project by Brilliant Labs. Purchased this and seems like might have treated it, or its case, a bit roughly. Decided that it was worth repurchasing this while waiting for their Frame iteration to come in. After a few hiccups with resetting Monocle via its Noa app on a few devices, am back at letting it be something of an assistant in training.

A preface: there’s almost nothing Monocle does out of the box except to connect to Noa through your OpenAI and Stability.AI API codes. You’d have to build the apps or connections to support what you want to do. This is definitely a “developer’s toy.” And one needs to know MicroPython in order to push that bit. Am using that to get back into Python a bit… and that was stuttered when the previous one stopped working.

What next? Well, am hoping to just get enough of the basics in mind so that am clear to what working with OpenAI and similar might look like. Today. It’s playing the role of answering a few questions (one of which has its answer staring at my right eye). Later, there might be some better HUD approaches alongside other devices (that whole “constellation of devices” approach comes in here.

If there’s some advancement with a local LLM project, am hopeful that it can be a part of that as well. The way forward for small businesses will be to do a hybrid of connected and local LLMs, and hopefully a decent audit trail for record keeping. This way of working seems to be the better part of “learn how this tech works and what gaps remain” for some.

Beyond that, the hope is to not break things, again. There’s a shape of this tech which is reparable and buildable - but also parts which need a kind of durability for a different kind of sustainability. It might be akin to the model Brilliant Labs is putting forth, it might not. But, it’s worth pushing forward to see what happens. It’s only a failure if one quits, right?