I’m not opposed to using AI as an aid, and my reply was not meant to say your suggestion was bad.
It’s not a stupid question if you are learning but still have things to learn. That’s just normal, and in this case discussing the issue at hand with someone more experienced will take you to the next level. It may even be rewarding for the expert, knowing that they invest their time making someone else an expert.
Remember that the AI is a stochastic parrot. It neither hallucinates nor knows anything, it just produces plausible sentences with words likely to occur close to the words of the prompt. And then words likely to be close to the words it has already chosen as a “reply”. It’s a coincidence when the collection of probable words conveys a fact, just as much as it is a coincidence when it confidently conveys a non-fact.
I think the person using the AI must also take full responsibility for reviewing and fact-checking its “replies”. This often seems to take at least as much time as doing some web searches and digesting the results, and is probably not rewarding for the expert.
Sorry, didn’t mean your comment in any negative way. 100% agree what you wrote. Always try to fact check AI answers. Best one I have heard when asked from AI how to increase/lower (can’t remember which) blood pressure with tricks in home and the answer was smoking. I mean in theory correct but wouldn’t recommend
Well, quite many does it like that and I would guess ~90% of people could probably change for example brake pads with help from youtube. Big portion even bigger components, in the end it is not so super hard if you are willing to spend the time and learn, isn’t it?
I got your point, I just think you are kind of wrong. If we use this car brake example, lets imagine AI is this mechanic who don’t have diploma or any certificate. He can do some kind of solution for the problem but you shouldn’t blindly trust that solution. You should still spend time to understand at least how to test that well and see that there aren’t any duck tape patents which might fail soon.
For sure someone who is as experienced as Michal can code better and faster than anybody without knowledge with AI. But if you can’t pay someone to do it, people probably don’t have time to gain similar level of knowledge for this first app they are wishing, is it best to just forget it? Or use tools which can help you make the app, or at least help determine if it is even possible? Sure it will need time, sure you might still need some help from others, but if you have the time why shouldn’t people try to do things by themselves? Sure it can fail, but so what? They hopefully learned something on the way
I have no idea how much work it takes to program an app. I thought it might be possible to take the Android app and, with minor adjustments, make it work on Sailfish. That would, of course, be more affordable than developing the app from scratch.
Bad example. You need a cylinder pusher clamp to change brake pads, or lots of manual skill (which you cannot get on Youtube) and curses (which you can).
Depends of car model? I have only experience of a little older cars where it has been pretty simple. Would be interesting to know if they have made that harder and why for newer cars but that is already quite off topic. Need to look it up
just a side note: since the android app is closed source, that is impossible. (if it was open source, it would be merely very difficult. it would probably be easier to make a native sf app from scratch, using the android source code for the api or protocol)
Hello,
is there still interest in a native Bluetooth (BLE) app for AntBMS (battery management systems)?
I would be available to work on it but need a volunteer beta-tester. If anyone happens to own such a device, please send me a PM.
Ok, so development is mostly complete and GUI is pretty much ready. Is the original author of this thread @rw007 still interested? Or maybe volunteers here who can help testing on real hardware?
In general: thank you for the development! Unfortunately I cannot help here, as I don’t need this. It isapity, that the original requester seems to lost his interest…
Hi, nice to know we have this tool.
Till now, I avoid all equipment requiring an Android app.
But knowing such tools exist is opening horizons.
This may take more time than we would wish, when developing an app, but I think users will arrive to your app.
I know basic BMSes, more from the soldering iron pov. But not connected ones.
What equipment type/brands is concerned here? (I mean, are there some other brands using the protocols/names you cited?)
Thanks
20 characters
