It's really that simple, and that is not a romanticized view of technology. Indeed, I'd argue that this is the (reasonable) assumption that most naive internet users make about their data.
Saying it a different way, if javajosh were to concede that your definition of privacy is more useful and concede that you are making a clearer case for how such things developed historically, how does he gain some legal breathing room for his remotely stored documents?
It wouldn't be incoherent at all for a 28th amendment to address various strands of privacy concerns that have arisen over the years. But there needs to be some thought into the design of such an amendment, because it wouldn't be an easy set of analogies from existing protections.
And my normative claim is that this position is totally, completely, batshit insane.
> The data in my drafts folder is exactly the same as data in a paper journal that I have in my house, and is protected by exactly the same law, the 4th amendment.
The use of "is protected" versus "should be protected" seems to me to be inviting an informative discussion, not a normative one.
As an aside, I'm always surprised by how often people on HN talk about "should" versus "is." That's very weird for the engineer in me. You can never make progress in a normative discussion, at best you can boil the disagreement down to a disagreement in principle and leave it at that. E.g. I don't trust the government less than I do Google, Facebook, etc. If I'm willing to write something in my gmail, where a Googler can see it, I'm okay with the government seeing it. You almost certainly have a different perception of privacy and trust. A normative discussion on the subject is thus futile--who is "right" about what who and how much to trust private companies versus the government?
Perhaps we make this distinction because it's an important one. It always surprises me when an engineer confuses the two. "But the courts say that the gov't can access your data if it's not on your property," is NOT a counter argument to the statement "The 4th amendment should extend to data." The conversation cannot move forward unless both sides understand the difference between "should" and "is".
This is what I refer to as the "romanticized view" of technology. Your gmail draft is not like the paper journal you have in your house. It's Google's data on its hardware that its engineers have access to (in clear text!). You want to construct this metaphor, where the "physical location doesn't matter", but that's not the underlying nature of the system.
You say that "representation" shouldn't matter, but you're making the opposite argument. You want different rules for digital representations versus physical ones. The rule right now is that once the information, represented as molecules of ink on fibers of paper, is in someone else's possession, it's not your information anymore. Well at the physical level, your gmail drafts are little flipped magnetic domains on a hard drive platter in a Google data center. If you tried to enter that data center, you'd be thrown out for trespassing. But you think that in this case, the law should construct a metaphor: those bits are "private" even though you don't have possession of them or ownership of the medium on which they reside.
Thought experiment: if I chisel my diary into a rock slab and mail it to Google, do you agree that it's their data now? What if I write it to a magnetic hard drive and mail it to them? No difference, right? So why should it suddenly be different if I send the bits over the internet for Google to write to its own hard drive instead of mailing them a hard drive myself?