I obviously don't know if it's happening or not (deleted my account years ago) but it isn't hard.
There's also the fact that Facebook has gone on the record specifically disavowing this conspiracy. If it were true, they wouldn't so unequivocally deny it. [0] There's far more harm for them deceiving customers than there is in them admitting to a creepy feature.
Additionally, everyone who has promulgated this conspiracy has been fairly suspect. None of them make any effort to analyze the alternative reasons.
But, like I said, I have $1,000 for whoever can prove this is happening. It really shouldn't be hard. If this conspiracy works the way people claim it does, you should be able to mutter about any random product to your phone and have Facebook start showing you ads for it.
[0] http://www.theverge.com/2016/6/3/11854860/facebook-smartphon...
That's a strange conclusion to come to. I think it was Shakespeare addressed that very thing in Hamlet.
>There's far more harm for them deceiving customers than there is in them admitting to a creepy feature.
If they've already deceived, then they certainly wouldn't admit it. If they stopped doing it, they could deny it was currently happening and not be lying.
>I have $1,000 for whoever can prove this is happening.
That number pretty much ensures anyone who has the ability to find out wouldn't bother.
Personally, I think they might have toyed with it but abandoned it. The trick isn't collecting the data without getting caught, that's simple (as earlier described). The trick is taking action on the data and not getting noticed.
If you use the conversation data as a weight to only be added to non-conversation data, you could do it completely unnoticed. In other words, gather conversation data but only apply it to data collected in other, less controversial means. So +1 for posting about Colgate, +1 for talking about it only if there is already +1 from the posting.
Why? I posit that a good portion of the customer fallout from this functionality has already happened since so many people believe the conspiracy.
By lying about it, they add a whole layer of difficulty. They'd also open themselves up to lawsuits: companies can't just record your audio and then lie about recording it.
> That number pretty much ensures anyone who has the ability to find out wouldn't bother.
Why? It shouldn't take any sophisticated technical analysis at all.
Ideally we'd have actual network activity proving it, but I'd settle for someone talking to their phone about a product which I choose for them (ie. not something they'd already have on their consciousness) and having ads show up for it. If this feature works how people claim it does, you should be able to get ads on demand for anything you talk about.
The quote is "The lady doth protest too much, methinks." I'm not sure if you are familiar with it but it means that when people vehemently protest something it usually means they did it.
I actually read the link you supplied and I don't think they are currently doing it based on what they said, but they might have done it in the past and stopped once people started noticing. They could be sending it to a third party, which would make their denial true. They might not be doing it at all. Who knows?
>companies can't just record your audio and then lie about recording it.
Depends on what people agreed to in the terms and also what wording they used to deny it. I mean they denied they gave NSA direct access to their servers, but NSA had direct access to their servers.
>Why? It shouldn't take any sophisticated technical analysis at all.
It can be easily obfuscated using compression, encryption and when they sent the data. I touched on this in another post on this thread.
They could gather the data and only apply it as a weight to an already collected data point that isn't related to the microphone. I posted this elsewhere also.
Aren't conspiracy theories fun?
I'm quite familiar with the quote. Except that in practice, covering up a negative action is often exactly what lands people into hot water (see Watergate, Clinton, etc.).
Facebook strenuously denies this conspiracy. [0] There are in fact laws governing consent for recording.
Of course, it's impossible to prove that Facebook has never listened in on people's conversations, but it's a very suspect claim to say they are today.
> I mean they denied they gave NSA direct access to their servers, but NSA had direct access to their servers.
They were legally obligated to deny that, so the exact opposite of the situation today (you're legally obligated to disclose if you're recording someone).
> They could gather the data and only apply it as a weight to an already collected data point that isn't related to the microphone.
Perhaps, but that's an entirely different claim than what the conspiracy theorists are claiming.
[0] http://newsroom.fb.com/news/h/facebook-does-not-use-your-pho...
You are mixing up testifying under oath versus a PR release. Companies lie to you all day long in PR releases. If FB were to go under oath and state that, it would be more believable, but not completely, see James Clapper.
I'm just pointing out how it's technically and orginizationally feasible to do this undetected. The only thing stopping FB from doing this is ethics and possibly backlash.
That said, it'll never make the messages smaller
I then encrypted the entire message you replied to and it was only 623 bytes. The original was 449 bytes.
Also, you can further obfuscate it by playing with the compression levels and algorithms.
The point being, the difference of 255 bytes is hardly enough to draw attention and hardly enough to hit data caps or affect battery life.