> The 'DeFi' systems you mention are not in control of vast sums of money, like the exchanges are. People want to use the exchanges and solving the problem by t

> The 'DeFi' systems you mention are not in control of vast sums of money, like the exchanges are. People want to use the exchanges and solving the problem by telling people to do other things isn't effective. Uniswap alone has settled over a trillion USD. The current DeFi ecosystem on Ethereum (even now in the current market) has tens of billions of in value locked. I do not know what would be vast sums of money, it's still small in comparison with traditional finance but the technology is a few years old. > This seems to me to be a problem with society and not with technology, and trying to apply technological solutions to problems that have existed and will exist as long as people exist, seems to be arrogant and ineffective. Indeed, the fundamental problems are societal and human in nature. The point is minimizing the surface area where that human intervention can cause damage. If your argument held any merit we might as well get rid of double entry accounting. After all, it's just trying to apply a technological solution to human problems. https://en.wikipedia.org/wiki/Double-entry_bookkeeping https://en.wikipedia.org/wiki/Double-entry_bookkeeping > Since we have been iteratively working on financial systems for a lot longer, it seems practical to work to progress that way instead of wiping everything and starting from scratch, where we are reinventing the wheel and all of the tragic consequences along with that. It's not reinventing the wheel. It's building a stronger foundation. I'm going to assume your field of expertise is CS, if it's not, hopefully it will be useful for someone else. Imagine the same argument brought to progress in programming language design. What's the point of Rust? We are just reinventing the wheel, considering how much has been built in unsafe languages like C or C++ we might as well keep iteratively building on them. Humans are still making programming mistakes. We don't need no memory safety in programming languages.