AI today is even more than an intern.
The event that brought me into today was an “Oh, shit” moment. I asked AI to do something, and not only did it do it, but it thought through the process, created the tools to do it, and then talked to me like a human.
Today, I can give it tasks, ask it to do research, and ask it to create requirements and then design, build, and test operational code. Here’s an example:
“We need to build a new automated test framework. The system must be trustworthy, run lights-out, analyze its results and publish them, self-heal when systems crash, and run the testing as fast as possible. Do the research reagrding our systems and industry best practices, generate a requirements document, and design an system architecture. For each functional item, decide whether it should be realized through COTS or custom-built tools. When you have the design, generate a proposal for me with pros and cons so we can discuss before we take the next steps.”
It will access the internet. It will use my computer and credentials to access all of our internal knowledge base, servers, and repos. It will use its reasoning skills. It might even prototype some parts along the way. And once we’ve discussed and agreed upon its design, it will do the implementation, testing, and debugging.
It’s moved up the chain from intern to a contractor brought in for a specific purpose or a fully sentient junior employee, and it’s moving toward mid-level.
How about the future?
You’re going to hire AI agents just like you would any other team member. Agents will take direction and talk to other agents. They’ll propose solutions. They’ll find problems. They’ll produce results. They’ll find ways to do it better.
All on their own.
The AI Agent Workforce is coming.
How will you get AI working for you?
Discover more from johnmaconline
Subscribe to get the latest posts sent to your email.