There are middle grounds here. Sovereignty, like security, is not binary, it's a spectrum.
You could use a scheme where two of three signatures allow transactions : you, your 'bank' and a notary. Or have such schemes for anything above a value. Or allow transactions by and to a list of whitelisted addresses after a certain date or event (such as a missed canary).
> You could use a scheme where two of three signatures allow transactions
So now social engineering hits and someone convinces my ‘bank’ and notary to allow the transaction to a third party. Now my life savings are gone forever.
In my real bank (no quotes) you have regulation. Banks are liable for fraud. Blockchains aren’t.
Now you could have pre assigned addresses that the notary and bank are allowed to transfer to, but now you have to think about edge cases. Every time i update my will I have to hire a blockchain engineer to update my smart contract and then hire another firm to audit it for errors so a bug doesn’t drain my account. Huge headache and much more expensive compared to the status quo.
> Or allow transactions by and to a list of whitelisted addresses after a certain date or event (such as a missed canary).
whats triggering the canary? How can I trust that it will be accurate?
Also, my spouse can change or die. If my spouse and I died at the same time and my parent’s should get the money if they are still alive. now what happens? You have to map out hundreds of edge cases, like every permutation of my relatives wallets. Then I have to trust that the blockchain can tell these people are alive.
Current California State law takes all of that into account. It handles the edge cases in a way that code does not. Probate court is expensive but its better than everything going to zero.
You are dismissing the examples for being too simple. They are simple because they are examples.
My point is not that these three examples "solve all issues we have with Authority/Govt/BigFinance". Nor to prove that "Bitcoin Fixes this".
My point is to show that crypto-currencies, by design, are programmable, and this allows a lot of use cases to move from Authority/Govt/BigFinance to DIY.
Not all cases, not everywhere, not for everyone. But in a spectrum.
To clarify with your refutations:
> Now my life savings are gone forever.
Which is why you probably want "your life savings" in a secure, goverened, trusted environment. For many people Blockchain or DIY is out of the question for "my life savings". But that leaves a giant spectrum of other use-cases. From checking accounts to saving-for-a-friends-present. Use your imagination.
> Then I have to trust that the blockchain can tell these people are alive
You are dismissing a giant set of use-cases on one that will, indeed, probably never work.
I never mentioned that this the ultimate and only replacement for Deed. I deliberately did not, because I think this is a bad use-case for Blockchain for the reasons you mention and much more.
But this could be a "here's some money for when you turn 18" or "bonuses for managers which can only get extracted after a year", a "tip jar that pays out every month" etc.
My point was that trust is a spectrum, and not binary. Yet you take the most extreme outliers thinkable, refute those and then conclude that none of the cases in the broad spectrum can ever work. They can. And Blockchain is one of the tools. As are trust-funds, banks, payment-APIs, courts, notaries and so on.
If you notice the phrasing, all the crypto arguments are in the realm "it could be...". The major difference is that all the legislation, be it good or bad, "is". Which makes crypto very convenient to paint imaginative rosy futures, and actual legislation very convenient to point out actual flaws and dark patterns.
You could use a scheme where two of three signatures allow transactions : you, your 'bank' and a notary. Or have such schemes for anything above a value. Or allow transactions by and to a list of whitelisted addresses after a certain date or event (such as a missed canary).