Agree with you on that, it was my concern too, but the way I think about it is access to information, the goal is not to provide hallucinations with a straight face (aka GPT), but rather use it as a way to extract necessary information fast. For instance, I have a built-in RAG that reads of growing collection on books on medical, survival, etc. (https://github.com/dmitry-grechko/waycore-knowledge) that AI agent is using to answer questions. Moreover, it has a built-in safety loop to always inform users on the accuracy of the information, but also if the information request has an impact on health & safety, it will warn users about it too.
So, I certainly see the inherited risk and problems, but mostly think about it as a means of information extraction
Thanks, haven't seen it before, but love the idea of it. It certainly takes the concept a lot futher that I intended with this project. My goal is far is to relyt on well-supported open source tools and frameworks, and give both the software and hardware flexibility for people to create what they need. Think of it like FlipperZero, but for outdoors
Love that I am not alone in this thinking! I test it on embedded systems (ex now Raspberry Pi 5) coupled with ESP32 as a sidecar. So, the core OS is running on the Pi and is reading data off the ESP32-connected sensors, basically.
wonderful... this is something I hope would come up going forward. I see this a little in the japanese electronic dictionaries still being developed and released
Will the screen be daylight viewable? (and no, trying to out-bright the sun on a battery-powered device is not a valid answer)
E-ink or transflective LCD or maybe the modified LED used by the Daylight Computer folks.
Agree that AI needs to go as not reliable enough for life-death situations.
Good point, I’ll be testing both. e-ink is great from power management standpoint, but I want to see how some apps will be rendering (ex maps)
> agentic AI
Yeah I don't want LLMs near anything life or death, where a hallucination can kill, thank you very much.
Agree with you on that, it was my concern too, but the way I think about it is access to information, the goal is not to provide hallucinations with a straight face (aka GPT), but rather use it as a way to extract necessary information fast. For instance, I have a built-in RAG that reads of growing collection on books on medical, survival, etc. (https://github.com/dmitry-grechko/waycore-knowledge) that AI agent is using to answer questions. Moreover, it has a built-in safety loop to always inform users on the accuracy of the information, but also if the information request has an impact on health & safety, it will warn users about it too.
So, I certainly see the inherited risk and problems, but mostly think about it as a means of information extraction
Putting the lookup in the AI means it can hallucinate the lookup. Putting the assesment of risk in the AI means it can fail on the assessment.
Please reconsider using a full text search index instead.
This reminds me of https://radiant.computer/
Any relation to the Daylight Computer folks?
https://daylightcomputer.com/
Not on my side, but it looks awesome. +1 to the e-ink point above.
Thanks, haven't seen it before, but love the idea of it. It certainly takes the concept a lot futher that I intended with this project. My goal is far is to relyt on well-supported open source tools and frameworks, and give both the software and hardware flexibility for people to create what they need. Think of it like FlipperZero, but for outdoors
Very similar to something I have been building this year. Do you already have hardware side ideas anywhere?
Love that I am not alone in this thinking! I test it on embedded systems (ex now Raspberry Pi 5) coupled with ESP32 as a sidecar. So, the core OS is running on the Pi and is reading data off the ESP32-connected sensors, basically.
wonderful... this is something I hope would come up going forward. I see this a little in the japanese electronic dictionaries still being developed and released