[LA_Banker]

This is interesting because while Apple's (comparative) dedication to privacy is endearing, it's a long-term existential threat.

Google knows all about me and its assistant is, usually, great. Amazon has troves of data on what I buy, and I get to yell at Alexa to order more TP as soon as I see we're on the last roll.

Apple knows much less about me and, while I'm still an Apple fan and am tied to iPhones/Macs thanks to iMessage, Siri stinks as a result.

If voice assistants based on machine learning (specifically, personalized voice assistants) are the next big thing, Apple's privacy ethos will separate it from its major tech competitors – either in a great way, or a very negative way.

[jswny]

Totally agree. Lots of commentators here on HN love to get angry about companies like Google collecting so much of their data, which is perfectly justified. However, you have to be willing to accept the consequences of privacy policy like that, which is the kind that Apple somehow still carries out stringently. The consequences are that any service or product that relies on data, ML, AI etc. aren't going to work well coming from a privacy conscious company like Apple, at least not nearly as well as the products from the companies like Google which don't value privacy as much. You can't complain about Google taking your data but also complain about Siri being a load of garbage, you can't have it both ways (and I see lots of people here trying to have it both ways).

It will be interesting to see if Apple holds its ground on privacy with the increase of AI/ML driven features and products. Coupled with Apple's closed culture which discourages open research (although they have been improving this), their concern for privacy could put them well behind other companies in this space. Depending on how you look at privacy vs. product, this could be a good thing or bad thing.

[naravara]

Aside from recommendation engines and targeted advertising, what does this deep knowledge of everywhere I've been and everything I've done do to dramatically improve the experience of using the AI?

I think it makes Apple slower at making their assistant good at parsing what you're saying and returning an answer, but that's a problem that benefits from crunching reams of data in general, not so much knowing everything about you personally.

For one thing, a virtual assistant that works even when I don't have an active internet connection seems like a perk in itself no? The Siri approach is closer to being there than the Alexa/Google approach.

[spiderfarmer]

What can Google do in the cloud that Apple can't do on the device?

[kingbirdy]

Apple can only look at you to learn - Google can look at you, and a million other people like you

[jug]

I think many Apple users would be happy if only Siri could answer factual questions objectively and follow a conversation, being happy with a minimum of privacy collection like current location, name of spouse for messaging, etc.

While some may prefer a knowledge about you or you via people like you, as in, "Please recommend me a great movie", Siri is currently not in a stage where she may tell you stuff like "How do I mix a White Russian?" Google Assistant give me 6 steps with a photo and a follow up question "What about Screwdriver?" Siri gets utterly wrecked in the knowledge based questions.

I think the main problem is not privacy-related, but knowledge base-related. Google is building upon a fricking huge search engine via a knowledge graph and the sky is the limit for how well an AI can do. Apple is building on what, a shut down Ping social network, Apple Music listening habits, Wolfram Alpha hopes if all else fails, and a sparse Bing Search API if that failed too?

[naravara]

>While some may prefer a knowledge about you or you via people like you, as in, "Please recommend me a great movie". . .

Not to mention that Apple positioned itself as a brand for people who "Think Different." I don't think people who got taken in by that messaging would be attracted to the prospect of services that can more efficiently pigeonhole them.

[naravara]

I thought Apple figured out a way to create unique IDs and profiles that can't be traced back to the individual person? So they can still find people like you for analytical purposes to tune their models.

[sheeshkebab]

Probably can't get all the metrics and data needed to actually tune/make it work.

I periodically delete all my history from google servers, thanks to their privacy tool, and google gets just a dumb for me as siri.

[Eridrus]

Speech recognition and natural language understanding.

If you watched the HomePod reveal you'd know that it sends your speech to the cloud for understanding. Which seems like a pretty clear admission that this can't all be done on-device easily.

[dr1337]

The simplest example would be speech recognition. With the exception of Keyword Spotting Systems like Hey Siri or OK Google, it's currently practically impossible to implement larger vocabulary speech recognition on device.

[gervase]

Google will happily permit you to download trained models of <100MB to your phone via Google Translate, which permit not only offline vocabulary recognition of a remarkably wide range, but will also then translate that input into another language (also offline).

I believe the restrictions imposed by keyword-spotting have more to do with the always-listening and/or power-efficient nature of the task, rather than the restrictions on the legibility of offline speech recognition.

[dragonwriter]

I'm pretty sure larger vocabulary local speech recognition on devices with less computing power than modern smartphones has existed (products for continuous speech recognition on PC go back to at least Dragon Naturally Speaking in 1997.)