Chainbase Phase 1 ~ The Issue of AI Consuming On-Chain Data
Chainbase, when reversed, is 'Chain Foundation', a name that exudes ambition. So what is this foundation (infrastructure)? Chainbase believes it is data standards that solve the interoperability issues caused by different data structures across chains. And why is interoperability needed? Because the data on the chain is ultimately meant to be consumed by AI, and if the standards are different, efficiency will be low.
Chainbase serves as infrastructure, returning to economic theory; it is necessary and will definitely be profitable. Everyone knows we are infrastructure enthusiasts (no wonder Tencent is interested; they have the infrastructure enthusiast gene).
So how powerful is Chainbase really? Or rather, is the work of infrastructure in Web3 as profitable as infrastructure work in real life? Both infrastructure and standards feel centralized; is it appropriate to do centralized things in a decentralized space? With so many questions, let's discuss them one by one. In this episode, we will first talk about the matter of AI consuming data.
2024 is a crucial year for AI because a significant event caused AI valuations to plummet, proving that AI currently cannot produce consciousness. There are two main forces exploring consciousness: one believes that the human brain is the source of consciousness, investigating the chemical and electrical signals of the brain; this European team has spent hundreds of billions exploring for ten years with no results. The other speculates that consciousness is related to the scale of neurons; they increase the number of neurons in neural network systems exponentially using computers to find consciousness under natural language models. However, both have yielded nothing.
This has caused the imaginative space of AI to suddenly shift from potentially producing silicon-based life forms back to large natural language models. It's like going from Transformers back to Baidu and Google. Thus, the importance of AI consuming data may have significantly decreased; what can it do after consuming the data? The enhancement of functions seems to lack any singularity for now; a better tool is just the ceiling.
