AI Tools Evolution: From Web Scraping to MCP Integration
Explains the Model Context Protocol (MCP), an open standard that allows AI tools to communicate with external services in a predictable way.
One of the challenges of using a tool like ChatGPT or Claude is that their capabilities are limited. OpenAI introduced operator and andthropic launched computer use in the last few months. However, even these are limited. The challenge is that these tools are built to imitate how a human reacts with the web, but AI isn’t a human. So it makes more sense that machines talk to machines rather than machines pretending to be humans, talk to websites built for humans. This problem has solved a long time ago. When you open the weather app on your phone, your phone uses something called an API to get the information it needs from another computer on the internet. The great thing about this is that the format in which your phone requests the information and the way it gets that information back is known and predictable. When we want to tool like ChatGPT to go beyond a chat and start interacting with other services, generally we need to babysit what’s going on, continually guiding it to make sure it’s going to the right website and looking at the right information. However, there’s a new way of doing things that makes it much easier for AI tools to talk to other services in a known and predictable way. It’s called the model context protocol or MCP. Now, the name doesn’t really tell you much about what it is or how it works. However, MCP allows for services like the weather that I mentioned before to be made available to your AI chat tools. It’s an open standard and there are only a handful of tools that support MCP today. However, dozens and dozens of services are being built to support MCP and more every day. For example, there’s an MCP connection for Salesforce so you can ask your AI about customer information and another one for connecting to any database to ask questions about your data using natural language. MCP is gaining a lot of traction. It allows you to get your AI chats to do so much more in a predictable way without having to do the babysitting. If you want to know more about this comment below with your questions.