4 ms·
> So basically, you guys are re-inventing what tradfi had years ago, just with worthless(?) tokens instead of debt backed assets. Crypto has debt backed primit
by hailwren 4y ago
> So basically, you guys are re-inventing what tradfi had years ago, just with worthless(?) tokens instead of debt backed assets.
Crypto has debt backed primitives as well. You’ve linked one. AAVE is a lending protocol. If you don’t like AAVE, maybe you’re looking for DAI.
> but no way to cash out to fiat.
This is just a result of US KYC regulation. Like it or not, this isn’t a choice made by crypto.
> What's the value proposition here?
The value prop is, as the gp said, trustless composability. Crypto (post eth) is to money as the internet is to a mainframe. A kid in a garage can build a financial primitive on top of a “blue chip” and the “blue chip” builder doesn’t have to worry (vis a vis bugs, which have been a big problem but a lot of progress has been made in that area).
I’m not claiming crypto(I wish there was a word for crypto minus btc) has achieved its end state, but the goal and road map seem very clear. Composable, trust-less finance is a strong use case.
Aside — you really have to decide whether the price is important in crypto or not. Opponents often use phrases like “worthless tokens” and then claim they don’t care ab the price.
- chao- 4y agoI do not know much about sophisticated finance. Can you explain the meaning of this phrase, or sketch out an example, or provide links to assist my understanding? >A kid in a garage can build a financial primitive on top of a “blue chip” and the “blue chip” builder doesn’t have to worry (vis a vis bugs). I get that a "blue chip" is supposed to be a big brand (at least that's how I hear the term often used?) but what does it mean to "build a financial primitive" and how does it relate to this (presumably third-party) blue chip brand?
- hailwren 4y agoRight, so it turns out that there are a fair amount of operations you can perform on financial objects to change their properties. You can buy, sell, combine, or split their risk profiles in different manners to get properties that are useful in different scenarios. Mortgage tranches played a big role in 2008, Black-Scholes helps us describe the behavior of options contracts, and maybe the most accessible for someone not familiar w/ finance is the Ray Dalio McNugget story - https://www.cnbc.com/2018/05/03/how-ray-dalio-helped-launch-mcdonalds-chicken-mcnugget.html https://www.cnbc.com/2018/05/03/how-ray-dalio-helped-launch-... The EVM specifically gives you a Turing complete financial system and, hence, probably a complete expression of operations on financial structures.
- this_user 4y ago> The EVM specifically gives you a Turing complete financial system and, hence, probably a complete expression of operations on financial structures. Which is nice, but doesn't solve any problems that actually exist. That is the running theme with all of these crypto/blockchain/"defi" projects, because they are created by tech people who don't understand the domain. They just start building all of these things that seem useful to tech people, but completely miss what is actually important.
- hailwren 4y ago> because they are created by tech people who don't understand the domain. They just start building all of these things that seem useful to tech people, but completely miss what is actually important. This is just false. idk what to tell you, but there are plenty of major builders in the crypto space who are highly regarded in tradfi. See ie jump Even if it were true (it’s not), it’s also a thing that could be said about any product built by tech people and hasn’t really seemed to play out that way in most other places.
- splix 4y agoCan you clarify what is "solving a problem that actually exist" means? I just looked around and I'm not sure I have anything that solves an actual problem. Like the HN right now, it doesn't solve anything, right. In the morning I bought stuff on Amazon, but it also doesn't solve any real problem. Not sure AWS is solving anything since we can rent servers directly. I used Zoom for a call, but it solves nothing, I guess. Also used Github, but it's just a useless service because we can use plain Git. I can't really find a tech product I use that clearly "solves a problem". The only problem solver I can recall is my coffee maker, but it's an analog device. So I'm wondering, what it's the criteria to decided if something solves an actual problem or not?
- notamy 4y agoNot the original commenter, but here's my stab at it: > In the morning I bought stuff on Amazon, but it also doesn't solve any real problem. Sure it does. It solves the problem of "I need something, I have money, and I don't have the time/ability/car/... to go to a physical location to buy it." > Not sure AWS is solving anything since we can rent servers directly. AWS gives you you - The ability to spin up MANY classes of servers instantly, whereas traditional server renting is more limited ime. - A host of extra services on top of it that all integrate fairly well with each other - Paid support if you want/need - ... > I used Zoom for a call, but it solves nothing, I guess. People like to see each other, phone calls can have quality issues and dead zones, ... --- Now, when I read the parent comments, I see: > The EVM specifically gives you a Turing complete financial system and, hence, probably a complete expression of operations on financial structures. And while that *sounds* interesting, I have absolutely no IDEA what the utility of this is. With AWS, Zoom, Amazon, ... I can point to things that very very directly benefit both technical and non-technical users day-to-day. That's not to say that there ISN'T use to the EVM -- I'm sure there probably is! -- just that imo it's less-obvious.