This is riding on the backs of community and trying to bait users into making tools for them for free.
yes, They might make tools easy for the community. Yes, It gives a lot of utility to the general populace. This is supposed to usher in the next wave of Automation and AI usage. I believe It's the classic case of using a jackhammer as a toothpick.
You just had to have had a standardized API. Swear upon pains of death, that this absolutely mandatorily requires an LLM for things to work the way they do now? All we are doing right now is fancy string parsing. Something that was done and all figured out from yore ( think back to the days of lisp ).
Love this! I’m Harrison, an ex fine dining industry line cook. My stack "The Secret Ingredient" adapts hit restaurant recipes (mostly NYC and L.A.) for easy home cooking.
At present, without considering their future ambitions for expansion, I believe MCP and A2A are complementary partners rather than competitors, with a relationship analogous to the hierarchical division of labor between microservices frameworks and enterprise communication protocols.
The core objective of MCP is to standardize interactions between large language models (LLMs) and external tools/data sources, enabling decoupling of models from resources via unified protocols such as JSON-RPC to enhance flexibility and security in tool utilization. In contrast, A2A focuses on enabling cross-system collaboration and communication among intelligent agents (Agents), defining rules for task allocation, state synchronization, and data exchange to support multi-agent coordination in distributed environments.
Together, they form a complete technical stack spanning model-tool interaction and system-level collaboration: MCP emphasizes expanding LLM capabilities through standardized tool integration, while A2A prioritizes cross-system resource orchestration. This creates a complementary ecosystem, bridging the gap between foundational model functionality and scalable, interoperable system architectures.
> Companies ARE Agents that other Agents can interact with.
I can't help but think that services-based companies will finally face disruption in a way they haven't so far. I tend to think about this post quite a lot: https://www.danhock.co/p/service-marketplaces
> But if that happens, guess which player in the industry I am betting on :)
This is riding on the backs of community and trying to bait users into making tools for them for free.
yes, They might make tools easy for the community. Yes, It gives a lot of utility to the general populace. This is supposed to usher in the next wave of Automation and AI usage. I believe It's the classic case of using a jackhammer as a toothpick.
You just had to have had a standardized API. Swear upon pains of death, that this absolutely mandatorily requires an LLM for things to work the way they do now? All we are doing right now is fancy string parsing. Something that was done and all figured out from yore ( think back to the days of lisp ).
Love this! I’m Harrison, an ex fine dining industry line cook. My stack "The Secret Ingredient" adapts hit restaurant recipes (mostly NYC and L.A.) for easy home cooking.
check us out:
https://thesecretingredient.substack.com
At present, without considering their future ambitions for expansion, I believe MCP and A2A are complementary partners rather than competitors, with a relationship analogous to the hierarchical division of labor between microservices frameworks and enterprise communication protocols.
The core objective of MCP is to standardize interactions between large language models (LLMs) and external tools/data sources, enabling decoupling of models from resources via unified protocols such as JSON-RPC to enhance flexibility and security in tool utilization. In contrast, A2A focuses on enabling cross-system collaboration and communication among intelligent agents (Agents), defining rules for task allocation, state synchronization, and data exchange to support multi-agent coordination in distributed environments.
Together, they form a complete technical stack spanning model-tool interaction and system-level collaboration: MCP emphasizes expanding LLM capabilities through standardized tool integration, while A2A prioritizes cross-system resource orchestration. This creates a complementary ecosystem, bridging the gap between foundational model functionality and scalable, interoperable system architectures.
Great post, thanks for sharing Aurimas!
> Companies ARE Agents that other Agents can interact with.
I can't help but think that services-based companies will finally face disruption in a way they haven't so far. I tend to think about this post quite a lot: https://www.danhock.co/p/service-marketplaces
> But if that happens, guess which player in the industry I am betting on :)
Long GOOG? :)