Google doesn't "spill their beans" to third parties-- what Google actually sells is the opportunity for third parties to be included in the advertisements Google is targeting to their users.

Google has a strong incentive to not allow their aggregated user data to leave Google-- the behavioral data Google collects is the reason why Google is valuable; if they start shipping that data off to third parties, suddenly the third parties don't need Google anymore.

(Same with Facebook-- they're not "selling" your data; they're selling the opportunity to target you based on your data, but the data itself is too valuable to Facebook to sell.)

> Google doesn't "spill their beans" to third parties

Except the government who gets unfettered access. Not into conspiracy theories but Im definitely not a fan of this (and it goes for all social media and technology companies)

> Except the government who gets unfettered access.

Does you have a source for that?

There was this Snowden guy who leaked some documents about that...

Then it will be easy to provide a link.

>Does you have a source for that?

Yep. They all do it.

[1] http://news.trust.org/item/20161004170601-99f8c

You have put a link about Yahoo! and that link says nothing about Google or others doing the same

No, its safe to say all major tech company's have have NSL's, and are required to allow the government to search and request data, it would be foolish to think otherwise. [1]

[1] http://www.nytimes.com/2016/10/05/technology/subpoenas-and-g...

Are you aware that your new link doesn't talk about unfettered access?

Yes, but this is all we know about now, every day there is a new revelation about what the government has access too that we previously thought they didn't. I think it's naive to assume they don't have access to all of it.

Your claim is that the government have unfettered access to ALL the data.

Can you provide a link to your claim or it is just your opinion because there is no proof of that?

Evidence has been slowly being released for years, what more do you need. Here is my third example for you [1], after this Im done.

[1] http://www.washingtonpost.com/wp-srv/special/politics/prism-...

You still have not provided any evidence for you first claim that the government has total access to those companies. And your last link also doesn't say what you think it says.

Google (or Facebook) is the third party. The first two are you and your computer.

It really is the big problem at the moment with the cutting edge of AI.

ML relies on large data-sets and if anyone tried to release a personal device it simply wouldn't even work, let alone compete with the mass surveillance google/ms/amazon are bringing to bear.

Unless the state-of-the-art in AI suddenly morphs, we seem to be stuck between giving up our privacy or having vaguely intelligent AI.

I personally fall heavily on the privacy side of stuff, but I can see the intellectual and commercial appeal of pretending it doesn't matter in order to get there.

This is simply a matter of AI not being advanced enough at the moment. Take how a human learns: it acquires experiences and information over a long childhood, and from this "data set" is able to do millisecond predictions of future events based off completely new data.

What needs to happen is a company needs to come along and create AI that is trained off of generalized information... some kind of socially accepted public data set... then the trained core is sold as a seed to individuals who then feed it their personal data.

It'll be the equivalent of buying an AI "teenager" and slowly training them to be an "adult".

I thing the fallacy here is that a single person can't create a large enough dataset, while I content they can. Combined with tools like pocketsphinx I think its very doable to have a privacy concious AI system.

That's the thing.

As sensors become richer and the data becomes more valuable to the ML, consumers are becoming more aware of their privacy.

That means to get to a 'good seam' in the future instead of trawling through trash, you're going to have to convince millions of people their interests won't be affected.

That means in time there is an opportunity for a Google-killer with a different business model not based on using the raw data OR using by the use of intelligent agents. Google goes down because its stakeholders are contingent on getting to the raw data.

I think we just envisioned a highly anthropomorphized AI: essentially, a very smart and entirely obedient person to serve as the perfect aide. The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away.

> The Star Trek dream emerged before computing technology was very far advanced, and well before the idea of constantly mobile wireless communication. We thought our AIs would be small and physical, easy for a single person to entirely own and unable to remember more details than a human; instead we got unfamiliar algorithms run on machines far away.

Which is kind of funny, because even if it might be accessed by personal mobile devices, the Star Trek "library computer" AI was never "small and physical, easy for a single person to entirely own and unable to remember more details than a human", it was an aspect of a large server (or networked cluster, the actual architecture is somewhat vague) that was part of a capital ship or base, had access in the server/cluster to a library of very nearly all generally available knowledge and extensive personal information about both its users and about people with little direct connection (and could reach out across a galactic network to access additional remote information sources to handle requests).

"Unfamiliar algorithms run on machines far away" is much more like the source of the "Star Trek dream" than "small and physical, easy for a single person to entirely own and unable to remember more details than a human".

I think Star Trek was very prescient here as well: Security is terrible in the Star Trek universe. Nearly every antagonist is able to break into the computer within about a minute.

True! I suppose I was thinking of earlier space-age fantasy more than TNG. It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship.

> It was rather odd how easily people could look up details on crewmembers aboard the Enterprise, but I suppose the situation's different on an organized ship.

Not just crewmembers: TNG showed some of the broader implications (both useful and creepy) of the convenience-oriented panopticon, e.g., when Geordi used the Enterprise library computer to construct a simulation from data (including personality profiles) of Dr. Leah Brahms, who later meets him (and encounters the simulation.) (S306 "Booby Trap", S416 "Galaxy's Child")

That was a substantial plot hole where she had no idea someone was simulating her.

I suspect a huge part of the panopticon culture would be / is being informed that you're being peeped at. 99% of the time someone asking "Computer locate commander Riker" involved commander Riker knowing all about who's looking for him and why and having a substantial conversation with the requester.

I don't recall any plot along the lines of Deanna getting jealous and spamming the computer all night asking where Riker is and he better not be in that cute ensign's bedroom... Because it seems logical the computer would inform him each time and he would eventually tire of the interruption and nature would take its course, WRT his relationship with Deanna.

A better analogy would be technically I could walk up to the company president's office and stalk him, but culturally that is so not going to fly and I would have a lot of explaining to do. Merely using the computer instead of walking there in flesh isn't a major cultural shift.

However, every single "holodeck creeper" plot line involved the simulated attractive real world woman not knowing she's being simulated until the plot reached maximum spaghetti spilling cringe, which is one of the few Trek panopticon situations where people being spied on did NOT know they were being spied on, which seems very un-trek, although it made for some entertaining stories.

An alternative interpretation is I believe over the course of the series every attractive woman on the ship was simulated on the holodeck at least once by at least one lonely guy, and its possible that culturally they just got used to it, although I find that unlikely. People do get conditioned to become used to the weirdest things, so its not out of the realm of possibility. Possibly a culture of what happens in vegas stays in vegas develops and its just the sexism of the TV show that they never showed the women turning the tables on their fellow male crewmen on the holodeck.

Isn't it the case that Apple's attempts at similar AI assistance store much of the data on the user's iPhone?