Hacker News Reader

> And as you know Siri and Alexa are always listening to every word so they can respond to the wake word. I believe recently Amazon admitted to storing all these records IIRC (please correct me if I am mis-remembering it).

This is not entirely correct the way I understand it. While they're listening as in, the mic is active, the full text processing is not happening until the trigger word. There's a reason Siri is called Siri (distinctive pattern, easy to pick up before applying a stricter check). The issue with the recording was that the device thought it heard the trigger word and the mismatched sample was still uploaded.

What I don't believe is happening is actual background conversation processing by the assistants on purpose. (There are going to be tech slipups) Simply because the moment that's revealed, they'll get a regulatory ban hammer they really do not want. It would also chew through either your data or battery and be easy to notice.

I don't put that much faith in TVs though for example...

Given that I feel that Apple, Amazon and Google all have deals with a number of key marketing companies where they sell your marketing id based on keywords their devices "hear".

Your feelings notwithstanding, there is zero evidence of this that I've seen.

Thanks for the detailed answer. That’s fascinating.

To be clear , you’re suggesting that there’s a map of labels (diapers, cereal etc) to marketing ids so apple can still say they anonymize your data even though for all intent and purposes, they don’t?

Correct, this is how some data aggregators sell your purchase data without actually selling your receipts. You will have a number of marketing id's from different aggregators, but people like Amazon, eBay etc will aggregate from multiple sources and can track you across aggregators. And there are re-targeting and re-marketing companies that do this as a service.

It lets them say they never sell your data, but they do sell meta-data about you. Probably a little over simplified but you get the idea.

Wow! If you tell someone that you purchased #123 and they know that #123 is shampoo, how is that not the same exact thing as telling them that you bought shampoo? I’m shocked at the sliminess of this.

You can guess this product affinity with a list of previous purchases. Just take a look at the 'arules' R package (https://github.com/mhahsler/arules).

You don't need to identify the customer.

I did some work with an online supermarket some time ago and I found some interesting associations worth trying like Sliced Brioche Bread + Nutella :-)

Sadly I didn't find any association like 'beer and diapers' ;-)

I thought the listening for the wake word was done in hardware for Alexa?