Actually, the internet of things is a brilliant idea, conceptually

The practice is hindered by short-term profits. It doesn't have to be this way, and a great many number of DIYers and hobbyists are trying to change that

I don't know if open source and secure projects will have feature parity with commercial Swiss cheese offerings, but we never will unless we try

Follow

@cypnk counterpoint: why the hell does your toaster need to be connected to the internet?

@troubleMoney I don't have a reason for a toaster, but I do for a coffee machine:

I'm asleep in my winter cabin, snow falling outside. As dawn breaks over the treeline, I open my eyes and say "cabin, what time is it?" Cabin answers in dry monotone "It's 8:20AM. Would you like me to make you coffee?" I say, "yes, please" as I stumble out of my futon, gaze at the gentle snow. Slow hum of water beginning to boil in the background

@cypnk I think you may have just reinvented the teasmade there

@Ninjatrappeur @troubleMoney The reality of voice recognition is that "intelligence", if you want to call it that, is a great deal more than just voice recognition

The "cognition" in voice "recognition" still replies on machine learning for the most part, which is why Siri and Alexa don't seem as alienating or creepy as a Microsoft Text-to-speech

Whether we like it or not, that kind of response does require internet connectivity backed by a server farm somewhere

Just wish it was open source

@cypnk I guess we disagree on this one.

I have a PocketSphynx instance running on a local rpi exactly doing this kind of stuff (task needed to be triggered when my hands are full).

Even if using cloud-based voice recognition sounds great, I think we should aknowledge we all live on a very limited amount of energy resources. We should stop using them in an un-reasoned way ASAP. We already can see consequences emerging...

Machine learning is energetically really expensive. Should we use it for document searching and language translation? Yes, I think it definitively worth it. We cannot solve these problem without this kind of technology.

Should we use it to make gadgets like voice recognition to turn on light or making coffee? I think not. Just use the switch or basic voice recognition if you suffer from a disability preventing you from using the switch.

@Ninjatrappeur On turning devices on/off, I don't think that's what most people stick to

In real use cases, people ask their digital companion all sorts of other questions, and get at least somewhat useful answers, which is arguably why people had the services in their homes in the first place

I don't think Siri would be as successful if voice to notes or turning devices on/off was its only feature (we used clapping before)

So adoption relies on more, which needs some kind of cloud

@cypnk @Ninjatrappeur - Siri refuses to clap for me, but if Alexa could gain the skill to applaud it would be a real ego boost. Plus it wouldn't be so dark in here.

@cypnk @Ninjatrappeur @troubleMoney There is open source speech.rec that runs nicely on local hardware.

@jerephil @cypnk @Ninjatrappeur @troubleMoney mycroft.ai/

It's worth googling it rather than just going to the link because you'll find more of the stuff they have going on.

@drwho @cypnk @Ninjatrappeur @troubleMoney Thanks. Looks incredibly interesting at first glance. I’ll dig into it at a later date.

@cypnk @troubleMoney

507 internal coffee server error, filter not found

Please contact your coffee serve administrator

(I got a timed bedside coffee machine in about 1978, never remembered to prep it the night before, doubt I'd remember to prep an IoT one, 40 years on)

Sign in to participate in the conversation
Computer Fairies

Computer Fairies is a Mastodon instance that aims to be as queer, friendly and furry as possible. We welcome all kinds of computer fairies!