[lifeisstillgood]

I agree with the idea that the best protecting n for privacy is to make it expensive to snoop - it's worked for thousands of years so far.

My concern is how to make the benefits of this huge amount of data (medical information, travel etc) possible while charging the other use.

It's just - how - does one attach the metadata and enforce the lolicensing?

[astrocyte]

Therein lies the 'product' ... I'm working on several projects so I wasn't shy to put the general idea of this one out there... It's one of many things I feel passionately are going to be changed by disruptive paradigm shifts ahead.

That being said, you come up w/ the 'solution' and you'll have your 'product'... I already have a general architecture/approach outlined for myself. Create one and, from the looks of the interest here, you'll have an audience to buy it.

Also, with the scale and depth of hacks occurring around the world, you'll soon have an even broader audience.

If you get funded, look me up ;). I'll be looking for work in the coming months =P.

[lifeisstillgood]

I'm wondering if / how the identity problem, blockchain proof of physical ownership and ...

Well, firstly applications should stop thinking they are CoW - data is something that gets passed in and operated on.

Secondly data operations should be side effect free - and distributed, so that we can ask for processing- in many ways I would expect not to share my data but to accept payment in return for some slice of processing someone wants

This may work ok, and generally the things we want (social networks) will be fine with local data sharing - facebooks walled garden would have to be replaced with open protocol (ie AOL -> SMTP) and no idea how to achieve that.

So im not setting the world on fire here. But I think the shape of future is clear - just getting there ...

[awgneo]

I love your perspective on the current obsession with cloud computing. Hit me up sometime if you would like to talk more about this topic.

[Natanael_L]

Multiparty Computation, homomorphic encryption, Private Information Retrieval, etc...

Unfortunately they're not very efficient, and have limited security guarantees. But that's part of an approach to do it securely.

[mtgx]

All of which Google likely has no interest in researching.