Hacker News Reader

Comment by thefounder | Hacker News Reader

Well Apple can start by making Siri work locally/without internet and then I believe their stance.

Cook isn't arguing that companies should not be allowed to have access to data. He's arguing that there should be laws in place to prevent misuse and abuses of that data.

There's nothing inherently wrong with agreeing to give a company access to your data. But it's absolutely wrong that once you've handed your data over, that company can do whatever it wants with it, with no restrictions, as if they own it.

How is Google abusing the data collected in your context? You gave your consent when you setup your android device, right? You want Google voice or Siri? They surely benefit from all your data.

In the same way you cannot just sign a contract that is against the law, a company shouldn't be allowed to do whatever they want with your data, just because they made you click on an "Accept" button.

So basically consent to use your data is a lifetime contract. I think you should be able to delete(I don't mean disable) your data from the platform when you sever ties with a platform.

If they could I believe they would. They rather not be reliant on whoever is providing their speak recognition.

A limited subset of functionality should already be possible, especially if they can have the speech recognition trained specifically to the user.

I get the impression that this is already happening to a certain extent and they are testing the water. On newer/faster devices, you’ll notice Siri transcribing on the screen nearly instantaneously, faster than any server round trip, but with poor accuracy. Then a few seconds later once the server connection is established you’ll see some of the transcription change to be more accurate.

Voice Control does this, but it’s not particularly great. I doubt it’s seen significant updates in years.

> They rather not be reliant on whoever is providing their speak recognition.

Apple has their own speech recognition tech.

They don't have the dataset needed to train the speech recognition engine, nor do they seem to be willing to deploy some of that $250b in cash to hire engineers smart enough to make a model which only needs the local user's input for training.

It might be an issue with processing power. The new NPU might fit nicely in this case.

Apple has had stripped-down versions of this before on both iOS and macOS via Voice Control.

There's a big difference between recognizing "open the photos app, computer" and "what times is the avenger's showing today, Siri?"

How do you plan to get the Avengers showtimes without connecting to the internet?

You misunderstand what the internet connection is for. Right now most machine learning models run on server farms and your voice snppet is sent to the server, where the model processes it. They then send the interpretation back

"Local model" means the voice snippet is processed on your device. Never beamed off to a server farm.

The interpretation might be:

  Action: internetSearch
  QueryString: movie times for "Avengers" [near me || {userPreferences.movieTheater}]
  UseLocation: true

Which then kicks off whatever process Siri has for handling internet search actions

I believe the "hey Siri" wake up works locally. For the rest that seems complicated as many of the questions you ask Siri requires Siri to fetch something online to find the response

"hey Siri"

- I'm sorry Dave, I don't have a network connection, I cannot respond to you.

"But you just did."

- I'm sorry Dave, I didn't understand you.

Reply on news.ycombinator.com