Meet Your Future Team Members: AI Agents

If you’ve been following the discussion around AI, you’re familiar with the concept of AI agents.

AI agents can be understood as intelligent automation that operates independently, monitoring its environment and taking action without constant human input. Unlike traditional software requiring specific inputs to produce predictable outputs, AI agents can adapt to varying conditions and user needs.

AI agents can be based on various technologies, including Large Language Models. They can also be constructed using other AI technologies, such as rule-based systems, machine learning algorithms, and specialized models tailored to specific tasks.

An expert interview with Alexander De Ridder of SmythOS

On my new podcast series, the AI AEC Show, I talked with Alexander De Ridder, the Belgian founder of SmythOS, now a US citizen living in Texas. He’s been working with Machine Learning since 2008.

SmythOS is an operating system with an intuitive interface designed for building and managing AI agents.

“Unlike other technologies that require you to have an entire environment and all dependencies tied to it, SmythOS can run with a small executable, fully self-contained, super efficient,” Alexander explained.

Users can deploy agents on local servers and enterprise clouds or integrate them into existing systems without being vendor-locked.

SmythOS fosters a collaborative workspace where teams can create and share agents. Users can set up spaces for different departments—such as marketing, sales, and development—allowing team members to work together on agent creation and deployment.

Agents for AEC

AI agents fit well within the dynamic construction business environment. Because they have agency, they can automate repetitive tasks, analyze data in real time, and make or suggest decisions.

For example, an AI agent could monitor project timelines, alert team members to potential delays, and recommend adjustments to keep the project on track.

Alexander says an agent can delegate work and collaborate with other systems through an API. It can control a web browser or a computer. If it controls a drone, it can take pictures, process them with vision algorithms, and update databases of inventory or progress automatically.

Alexander envisions a future where trillions of interconnected agents operate across various sectors, including construction. These agents will enhance productivity and create new economic opportunities.

“They will be everywhere. And some agents are small. They are like your smart vacuum cleaner. Other agents are large, and they orchestrate an entire supply chain for a country.”

Listen to the interview to learn more about the exciting future of AI agents.