AI Is the Next Great Interop Layer

I had previously observed that humans are the great interop layer—we are the glue that fits together disparate processes and tools into usable systems. After using large language models, I’m becoming convinced that they can offload a large amount of the interop cost that currently falls to us. In a nut shell, AI can ‘do what I mean not what I say’ pretty darn well.

The latest LLM tools can interpret what we mean far better than before. Natural language interfaces opens up the possibility of sloppy interfaces—ones where not everything needs to be specified precisely. This allows for tolerance in the system making it easier for more things to fit together.

By contrast, APIs are largely a ‘do what I say’ interface. They require precise steps to complete an action. They require documentation to help a human understand how to use them. It takes creativity to implement them in ways that can solve their problem. Now we have LLMs that can figure out how to make API calls (Zapier NLAPI) and map different actions to different methods (langchain tools).