A buddy sent this to me thinking that I would like it.
I can understand why he might think that, but I don’t.
I hate the idea of Alexa being between me any my guns.
I don’t know if she’s going lock me out of my guns when I need them or rat my ass out to the cops or the social media mob.
Cops: “We’re here for your guns.”
Me: “Tragic boating accident.”
Alexa: “Check in the side board officer.” *Opens gun cabinet*
The truth is I don’t trust that bitch Alexa not to be a fucking snitch.
No, Alexa doesn’t need to know or have anything to do with my gun stash. Fuck that noise.
Yeah, I absolutely dislike this interface. Besides the hacking/security issues with this, no doubt Alexa software is not held to the standards of critical control systems, and I’m also not a fan of anything that requires power.
Well, if you ignore the reliability issues – like real-time Internet availability in a critical SHTF situation – the “ratting you out” problem can by bypassed by using a different key phrase, like “Open the pod bay doors” or “January suborbital denomination.” Alexa doesn’t need to know what the controls do, or what’s in the drawer, only that hearing a particular phrase calls for activating a particular control signal.
But the real security issue here is building this awesome concealment/security contrivance, and then showing it off on the ‘Net.
Somewhere there has to be a database of those phrases and associated actions. A warrant might well get that released, and then it’s just a matter of looking through the list.
And, yeah, I get your meaning re self-breached security…
“Alexa, we have a warrant. Unlock everything.”
… I have notions of setting up some smart-home stuff,.but using my own dang private server, and not relying on Internet connectivity for internal functions. (This seems to involve building my own video doorbell, because last I checked you couldn’t buy one that could be configured to work with your own equipment – they’re all tied to Somebody Else’s Service. While I’m at it, maybe I’ll make it look like a scaled-down HAL 9000 terminal.)
The home NAS we use, Synology, apparently has support for security camera feed recording. A basic web cam plus something like that, might do for a pseudo-smart doorbell. Perhaps use a home automation controller to trigger recording for 10 sec when the doorbell button is pushed?
Just spitballing here, but I think the pieces might be available.
The question is whether these devices send stuff out of your home, with or without telling you. If they are made in China, I certainly would not trust them. If made elsewhere, I still wouldn’t trust the company to do the right thing.
Now if things like this are built using open source code, then that’s a different matter.
I refuse to have a smart speaker in the house. Period.
The phones … Well, we disabled what we can but Google keeps adding new “features” so it’s like playing whack-a-mole.
Back in March, my google smart watch went off at 6am for no discernable reason. I picked it up and it was….a reminder to wash my hands. Google pushed that out to every single device.
Later that month, the most conservative governor Tennessee has had in my lifetime announced he was shutting down the state because they performed a surveillance operation on every smartphone user in the state and determined people were leaving their houses too much.
I have a regular watch now and a dumb phone with no Bluetooth.
I am waiting for a non-internet based voice control, it might take a decade before the technology can match amazon’s or google’s current server based services but eventually i believe I will be able to play with voice control automation without allowing a 3rd party to spy on me. Maybe by then the at home facial recognition cameras will be developed to the point they no longer need a off site server as well. I was really geeking out last year when I realized all the neat things I could do with current home automation tech but then I realized most of it would allow someone to spy on me.
That has existed since the 90s. I used it circa 2001 for very specific use case and it worked well, even over two-way radio. Today, there are software packages with AI that will work on a raspberry pi. You just have to be a nerd to set it up.
I’m pretty sure there are open-source options out there. They might be more sensitive to the speaker’s voice than the “cloud” offerings, but one might consider that more feature than bug for a custom system.
And… have you looked at the price of a refurbished server with three-years-ago tech? Some of them have gotten crazy cheap, and should have plenty of CPU power for home voice processing / face recognition use.
(I’ve been running in-house servers for various functions for many years now. Had to move mail and web services to a rented VPS a couple of years back, as I was preparing to depart Silicon Valley, since I’d be losing the static routable IP address I’d had at home since forever.)