Comment by PierredeFermat | Hacker News Reader

Comment by PierredeFermat | Hacker News Reader

Most, if not all, arguments on data ownership are arguments by analogy. This makes them ipso facto invalid. Using that same line of inductive reasoning, we could argue that we own the trash we generate, and that's not a reductio ad absurdum.

An encrypted 'digital locker' is a naive and unreal solution, at least until homomorphic encryption is mature. How would I request data I'm interested in from people without knowing if they have it in the first place? Issue a million requests? And if I'm issuing such requests, do people have to pay me now because it's data that I have generated?

We are agreeing on the terms and policies of all information services we use and generate data from. It's not like they took us by surprise, except for few Analytica cases of course. If we follow the data ownership argument, which has (re)surfaced mostly due to media attention on FB and the likes, such services will become dysfunctional right away.

Privacy isn't the weak argument, it's just not very well enforced yet mostly because most people don't care much. Ideally, if a company wants to sell your data, as part of their ToS/PP, they notify you. If you approve, you get a share of that sale (basically what Nukleosome is doing). If you don't, they just move on.

> We are agreeing on the terms and policies of all information services we use and generate data from.

If only that were true.

How is it not true?

To be charitable, maybe we all click Agree in the same way that we consent to being searched at the airport: there's no other choice, so it's a begrudging accept, not an enthusiastic accept.

I didn't claim it's an enthusiastic one. My claim is that all the sudden "privacy-is-not-enough" arguments make it seem like some breach happened without our agreement. To use your words, it's almost as if they're complaining about airport security/searches because we were never very explicitly warned about it.

If we don't do security checks or don't agree to some ToS, we'll be left out (of flights or online services). How many people would do that in exchange for 'ownership'?

Also, if you go to a bakery for couple months buying the same bread. Would you sue the bakery if the seller automatically knew what you're going to get, after two months of transactions, and wrapped your favorite bread for you with a smile (not the Amazon smile)? Is that also data you should own and encrypt?

Does your physician need your consent on every visit to unlock your health information in his brain and be able to follow up?

Imagine that this comment I'm writing here is locked into my very own encrypted vault and HN needs my consent every time someone wants to read it (although I used HN to write it!). I cannot imagine how would the discourse evolve this way.

So justifying 'ownership' as the future solution using invalid argumentation by analogy isn't solid. A more solid approach is to actually build products with well-enforced privacy and transparency rather than just theorize about it. Or even build and experiment with the 'ownership' proposition and see how would that play out.

It's not enough anymore because many people are just learning about how widely their data is shared around.

For example, most people that use a loyalty card at their grocery store has no idea that what they buy is being sold to Facebook. People have just discovered that Facebook is tracking them even if they have never had a Facebook account and never consented to anything from Facebook. Same goes for Four Square. Not many people have a Four Square account yet Four Square has built a giant location data empire. MasterCard is trading your purchase data. The list of examples like this is long.

There needs to be some regulation around the ownership and use of information by and about individuals. I don't have a relationship with Four Square so they shouldn't be allowed to hold any data about me.

I'm not advocating against that, quite the opposite! My work involves solving exactly this class of problems and good privacy ensures such behavior does not take place.

What I don't see is how changing the nomenclature to 'ownership' would solve all of this. It's like calling some statistical models AI models. You see my point?

Gee it's almost like this entire conversation is a semantic argument over the definition of "agreement".

You know, I think it is time that we argue about definitions. Many of them are starting to get loose.

I've come to believe that most disagreements online come down to definition, and could go to more interesting places if both parties got crisp with definitions early instead of arguing X and X' 10 replies deep only to then realize that X and X' are subtly different.

Reply on news.ycombinator.com