I got a bunch more motion sensors and I've been putting them around the house so I can program Home Assistant to do stuff like turn off lights in rooms that aren't being used, turn on lights when we're going downstairs with hands full of laundry, obviously just flipping a switch is easy but in some rooms the switch is on the opposite side of the wall and if I put a button on the entry-side then I don't trip over the cat, telling the water heater to get hot before we shower, having a $10 smart plug tell me when my 80's washing machine has finished, having a button to make the lights go cosy, y'know just nice helpful stuff that I do because I like making the house better for my spouse and kid and as I was putting up all these motion sensors I was thinking,

man,

there's NOTHING in this program to detect when it's being used for evil

And like there really should be? Like someone could legit fucking torment their family with this, this could be used as a tool of manipulation and control and fostering the abusive sort of dependence

And then I think of how me and spouse are moving away from google and onto like self-hosted stuff, nextcloud and searxng and xmpp and shit like that, and I can't help but think, *know* really, that right now there's someone controlling their intimate partner with those same technologies, like something that's supposed to be liberating you from corporate-style General Abuse is being leveraged towards a form of very focused abuse against you specifically by someone on whom you depend

I don't think enough software devs spend enough time thinking about how their projects can be used to hurt people

Has anyone smarter than me written any long posts thinking about this and what we can do about it

@ifixcoinops What are some things manufacturers can do to mitigate abuse?

I’m asking honestly. I don’t really have any ideas beyond a kill switch that basically disconnects the system from network access.

@drahardja yeah that's the question I was asking

@ifixcoinops @drahardja on the hardware side, some locally-visible indication that the device is being controlled remotely (e.g. blinking for a few seconds after a command) + some way to easily disable that remote control + local controls so you still have the basic non-smart functionality?

Though it's cheaper not to add all that stuff so the cheapest option at the big box hardware store will probably also be the most abuser-friendly :/

@ifixcoinops Oh I see. I think at a minimum having local control of everything, and having an ability to one-push disconnect network control is a prerequisite. It doesn’t have to be on every switch or device, only at some designated hub.