The Web3 Framework for Applications
What exactly does it meant to build in "web3"?
To date, many have defined web3 as building on blockchain infrastructure. They are able to offer web2 type experiences, but the underlying infrastructure isn't run and operated by a single centralized provider.
The problem is that nature of the infrastructure doesn't, itself, affect the application experience for users.
Making the case that the decentranlized infrastructure warrants the shift is sufficient is like saying someone building an application on serverless micro-services instead of a single monolithic is a tectonic shift in applications and markets.
It largely isn't.
The underlying change in the types of infrastructure doesn't fundamentally change things.
Typically, it is when technology shifts ushers in business model innovation that something new and exciting enters.
This was the case of the original internet -- when the original internet emerged, new business models around personalization , SaaS, and e-commerce could kick into gear. Businesses had to compete because there were new business models such as subscription and made-to-order.
Something similar must be in play for web3 to warranty both a new framework and a new set of markets and users.
Otherwise, if changing the backend alone is the change, it's not much different than if Sears took photos from its mail-order catalogue and put them on a webpage. Without new business models, the technology cannot fulfill its promise.
Decentralization of infrastructure, on its own, is not sufficient.
However, we can see that DeFi has introduced a new way of finance by making it trust minimized in terms of participation. Now anyone can earn fees as a liquidity provider.
DeFi ushered in the ability to own, via tokens, a piece of the protocol and govern its policies -- making ownership in infrastructrue meaningful for participants.
This ownership enabled new customer behaviors in meat space. Helium owners can participate in building out wireless towers and owning usage.
The second principle is trust minimization: which again is more of a technical concept that doesn't have direct appeal to customers.
But what this should enable is for customers to benefit from permissionlessness and composability of data and services.
For example, a membership service could have increasing benefits, not because there is a BD team, but because providers who meet the service requirements and brand constraints should be able to offer services that benefit both themselves and the customer.
The third is verifiability -- and this specifically should reduce the data surface area. This seems like something that will overturn the surveillance economy, but it will only work if there are incentives. People advocating for their own privacy just isn't enough.
Likely this will have to start to come from governments and policy-makers to guide large companies to adopt this new form of technology.
It would be a big bet for a start-up to start offering this, but if it aligned, it could be similar to how SOX drove alot of initial customers to emerging startups, similar for GDPR. None of those really did much honestly, because regulations unfortunately tend to be net-bad; so the idea case is if there is a market force.
What would it take for an application development platform to enable this?
It should enable the easy creation of a native token.
It should support the creation of incentive mechanisms via smart contracts or off-chain compute to calculate and execute reward systems based on that token ownership.
It should provide flexible application of off-chain computing services like zero knoweldge proofs or other cryptographic primitives.