Self-addressable data is a familiar idea. IPFS uses a hash of data as a data address. Addressable data is everywhere, from memory and filesystems to databases and blockchains. If we focus on structured data, we could make a step forward. VC has some obvious limitations and downsides - JSON-LD is too noisy for IOT devices and systems that require a compact data representation. CBOR-LD and JSON-LD-based credentials, even for the same data, will have different signatures, and it is no way to compare and compose this data together.
The solution for this problem is a serialization mechanism (https://datatracker.ietf.org/doc/html/draft-ssmith-cesr) that focuses on data tuples (key, values ) and has special cryptographic placeholders for a cryptographic digest of data. So cryptographic digest of data caller SAID (https://github.com/WebOfTrust/ietf-said)- self-addressable id and serialized data with identifier inside form SAD - self-addressable data. Now we have a crypto-powered lego-like building block of self-reference, Authentic data, or ADC - Authentic data containers. Now we have a composable, human-friendly, and machine-friendly solution. What kind of problem could it solve? JSON-LD is used to build data graphs and knowledge graphs; take a look at my video about VC (https://www.youtube.com/watch?v=owbkzvLhblk) and an excellent explanation of JSON-LD (https://www.youtube.com/watch?v=vioCbTo3C-4). The problem with JSON-LD is that some critical data links are not an integral part of the data itself, like a link to a context or a revocation - it opens a potential vector for security attacks. SAID could be composed of a complex data graph that could hold an authentic reference to data and schema. Another example is Verifiable presentation - with SAID, we could include VCs that could refer to each other and blindly confirm user binding and ownership. We could express a complex relationship between VCs