[nfin]

> At the moment, it's quite clear that current cryptocurrencies are solving problems nobody has (trustless transactions)

That’s an assumption, not a fact. And trustless transactions might not be the only problem that it tries to solve. What about predictable money supply. Trustless custody (instead of just “trustless transactions), … All these will not appeal to a lot more people today (but nobody and not many is very different, and that ratio can change with future technologies being built)

> how to restore keys when they lose them or they are stolen, and related how to tie those keys to real-world identity in a meaningful way, how to rollback fraud and punish grifters, etc... for most of these you need a trusted central authority and also trusted, verified identity.

The first part (how to restore keys when they lose them or they are stolen) does not necessarily mean that there is no decentralized solution. Social recovery (Shamir Secret Sharing + social recovery; or safer some multisig + social recovery) is being worked on.

The second part “how to tie those keys to real-world identity ” is much harder (specially if one values anonimity to avoid 1984 scenarios).

[lottin]

The money supply is not observable. Prices are observable. People care about price stability not about stability of the money supply.

[nfin]

price stability has it own set of problems.

If a good was regulated to have a stable price nobody would be incentivized to find a clever solution to solve a future crisis with high demand for that good

[lottin]

Price stability refers to the stability of the price level. It has nothing to do with changes in individual prices which are unrelated to changes in the price level.

[grey-area]

> What about predictable money supply?

It's clear that current cryptocurrencies are absolutely not solving that problem either. There's a fact if you like facts, consult the Bitcoin price.

> Trustless custody

Not clear to me that people outside the crypto bubble want this; people want trusted counterparties, not trustless obfuscated counterparties. It is IMO a solution looking for a problem.

I'd be a lot more sympathetic to this space if wasn't full of grifters and fraud. As it is I think the crypto experiment has irreversibly been tainted by that association (and by people losing lots of money), and I would not trust a 'trustless' solution from any of the current crypto companies or individuals.

> how to restore keys when they lose them or they are stolen

This is a v. hard problem, why make it 100x harder by insisting on decentralising the solution? And then 1000x harder by insisting on anonymity? Those may be properties of your chosen solution, which is I suspect why you're insisting they are necessary, but they are a bad design IMO - these are the fundamental design flaws of current cryptocurrencies.

Normal people don't keep backup keys on a second device etc etc, web of trust is a very old idea which has been tried quite a few times (see pgp for example, keybase for another corporate one), and you need a way for a normal person to prove they are who they say they are and regain access via courts or a central authority, take over inherited accounts etc. At some point these systems have to interface with the real world and real world authorities and laws/courts.

Just to propose alternative solutions to safely storing cryptographic keys (note those are useful for all sorts of things and unrelated to cryptocurrencies):

Corporations like Apple, Google could provide such a service, as they already own most of the infrastructure. There are obvious and significant downsides to this.

Enlightened governments could propose such an infrastructure of identity verification and private keys, there are obviously problems with that too, but it could be workable if you trust your government.

Utopian techno-geeks could also provide such an infrastructure, but somebody has to pay for it, and people fundamentally have to trust the people who create and run the system - that's a hard problem without financial incentives for the devs/maintainers. One example of an existing system is DNS and another is certificate authorities - both are not great but do work in the real world for their intended purpose.

I do believe at some point we'll come to solve this problem of digital identity and authentication because it is so fundamental, both for humans and corporate entities. I'm not sure we'll like the solution which ends up winning, and I certainly don't think cryptocurrencies are a contender.

[nfin]

how is the bitcoin price supposed to reflect the supply.

The supply is pretty much math.

Yes, the longterm price trend is an indicator of people valuing that fact (and other properties). Short term is governed a lot more by media, fear, …