3 ms·

> The problem is there is no coherent programme on how to counter the privacy-encroachment trend. IMO, there sort of is: FOSS I have my own "Alexa". It uses F

> The problem is there is no coherent programme on how to counter the privacy-encroachment trend. IMO, there sort of is: FOSS I have my own "Alexa". It uses FOSS, and recognizes simple commands, like: lights off. Play/pause music. Add something to the grocery list. Unfortunately, it's a lot less feature-full. I really want an Alexa (or the like), but I personally refuse to get one until I can run it on my own servers, and block it from getting on the internet. And unfortunately, that doesn't look like it's gong to happen anytime soon. I think the big problem is the inability to run these systems without "big data" behind them, and I'm not sure what's the answer to that gap.

lukeschlather 11y ago
What verifiably private options are there for speech-to-text? It seems like they'd have to be FOSS but most of the "FOSS" solutions I've seen still use a remote server for the actual processing.
- walterbell 11y ago
  It doesn't have to be FOSS, all you need is software designed to run offline, which can be network isolated, e.g. in a virtual machine. Dragon Naturally Speaking (from Nuance) in a Windows VM could perform the speech recognition and relay commands to a Linux VM. See also the http://www.voxforge.org http://www.voxforge.org effort to aggregate OSS acoustic models.
  - CaptSpify 11y ago
    > It doesn't have to be FOSS, all you need is software designed to run offline, which can be network isolated, e.g. in a virtual machine. It depends on who you ask, some people may only be willing to run FOSS stuff |;) But yeah, I've looked for a "offline: FOSS or not" solution, and haven't found much. EDIT: I can't reply to you because thread is too deep I'm assuming, but: yeah, it looks like it does. Unfortunately I probably won't try this because it requires windows, and I'm not really wanting to set up a VM/Wine to accomplish this. But you are correct.
    - walterbell 11y ago
      Dragon works offline, http://venturebeat.com/2014/08/22/nuance-dragon-dictate-4-is-nearly-word-perfect-review/ http://venturebeat.com/2014/08/22/nuance-dragon-dictate-4-is...
- CaptSpify 11y ago
  > What verifiably private options are there for speech-to-text? Unfortunately not a lot. That's why I said it was "sort of" and answer. I've tried http://julius.osdn.jp/en_index.php http://julius.osdn.jp/en_index.php and https://github.com/cmusphinx/pocketsphinx https://github.com/cmusphinx/pocketsphinx in the past, with very limited results. Though it's been a while and they may have gotten better.
dunkelheit 11y ago
FOSS is from another era. Back then Microsoft were the bad guys and FOSS was largely a response to their hegemony. Ubuntu bug #1 was "Microsoft has a majority market share" and it is kind of resolved now but increasingly irrelevant. Now it is a different kind of challenge. True, if you are a geek, you can use open-source software to cobble together some resemblance of popular cloud services but under your own control. There are even detailed guides on how to do it (see also: https://github.com/sovereign/sovereign https://github.com/sovereign/sovereign). But that's a lot of nuisance and in the end what you get is an inferior version.
- CaptSpify 11y ago
  > Now it is a different kind of challenge. Howso? I don't see it as any different, but instead of Microsoft, there are multiple organizations. > FOSS is from another era. I'm not sure what you mean by this. FOSS is still current, and quite active today. In fact, I'd say even moreso. > But that's a lot of nuisance and in the end what you get is an inferior version. I agree. That's what I do now, but it's not nearly as good. But what I was implying was that we have the potential to create a much better version, that's trustworthy.

lukeschlather 11y ago

What verifiably private options are there for speech-to-text? It seems like they'd have to be FOSS but most of the "FOSS" solutions I've seen still use a remote server for the actual processing.

walterbell 11y ago
It doesn't have to be FOSS, all you need is software designed to run offline, which can be network isolated, e.g. in a virtual machine. Dragon Naturally Speaking (from Nuance) in a Windows VM could perform the speech recognition and relay commands to a Linux VM. See also the http://www.voxforge.org http://www.voxforge.org effort to aggregate OSS acoustic models.
- CaptSpify 11y ago
  > It doesn't have to be FOSS, all you need is software designed to run offline, which can be network isolated, e.g. in a virtual machine. It depends on who you ask, some people may only be willing to run FOSS stuff |;) But yeah, I've looked for a "offline: FOSS or not" solution, and haven't found much. EDIT: I can't reply to you because thread is too deep I'm assuming, but: yeah, it looks like it does. Unfortunately I probably won't try this because it requires windows, and I'm not really wanting to set up a VM/Wine to accomplish this. But you are correct.
  - walterbell 11y ago
    Dragon works offline, http://venturebeat.com/2014/08/22/nuance-dragon-dictate-4-is-nearly-word-perfect-review/ http://venturebeat.com/2014/08/22/nuance-dragon-dictate-4-is...
CaptSpify 11y ago
> What verifiably private options are there for speech-to-text? Unfortunately not a lot. That's why I said it was "sort of" and answer. I've tried http://julius.osdn.jp/en_index.php http://julius.osdn.jp/en_index.php and https://github.com/cmusphinx/pocketsphinx https://github.com/cmusphinx/pocketsphinx in the past, with very limited results. Though it's been a while and they may have gotten better.

dunkelheit 11y ago

FOSS is from another era. Back then Microsoft were the bad guys and FOSS was largely a response to their hegemony. Ubuntu bug #1 was "Microsoft has a majority market share" and it is kind of resolved now but increasingly irrelevant. Now it is a different kind of challenge. True, if you are a geek, you can use open-source software to cobble together some resemblance of popular cloud services but under your own control. There are even detailed guides on how to do it (see also: https://github.com/sovereign/sovereign https://github.com/sovereign/sovereign). But that's a lot of nuisance and in the end what you get is an inferior version.

CaptSpify 11y ago
> Now it is a different kind of challenge. Howso? I don't see it as any different, but instead of Microsoft, there are multiple organizations. > FOSS is from another era. I'm not sure what you mean by this. FOSS is still current, and quite active today. In fact, I'd say even moreso. > But that's a lot of nuisance and in the end what you get is an inferior version. I agree. That's what I do now, but it's not nearly as good. But what I was implying was that we have the potential to create a much better version, that's trustworthy.