Thanks, Josh. As you say, that backend unification is worrisome. I'm pretty confident Whatsapp didn't read your messages before the Facebook acquisition, but now I'm sure it does.
I have tested it a couple of times; chatting with friends about random stuff, never speaking nor searching about it in any other way, and still receiving ads about that same random thing. There's no way they could have figured that out without reading my messages :( I did these tests with and without a native keyboard with the same result. I could be missing something...
Hmm, could be that they track behaviour before and after a conversation for both participants and then follow that on to whoever else they talk to. Eventually, someone follows up on something and it gets traced all the way back to you?
I'm less inclined to believe they read messages than I am to believe that the ML algorithms and data collection process is now good enough to make it seem that way. I actually think that the latter is more dangerous because it's less obvious how they do it and the potential for misuse is higher e.g. Cambridge Analytica.
Thanks, Josh. As you say, that backend unification is worrisome. I'm pretty confident Whatsapp didn't read your messages before the Facebook acquisition, but now I'm sure it does.
I have tested it a couple of times; chatting with friends about random stuff, never speaking nor searching about it in any other way, and still receiving ads about that same random thing. There's no way they could have figured that out without reading my messages :( I did these tests with and without a native keyboard with the same result. I could be missing something...
Hmm, could be that they track behaviour before and after a conversation for both participants and then follow that on to whoever else they talk to. Eventually, someone follows up on something and it gets traced all the way back to you?
I'm less inclined to believe they read messages than I am to believe that the ML algorithms and data collection process is now good enough to make it seem that way. I actually think that the latter is more dangerous because it's less obvious how they do it and the potential for misuse is higher e.g. Cambridge Analytica.
Really interesting thought 🤔