Jon Krohn believes that a system is only truly agentic if it operates autonomously rather than requiring constant prompts for every action.

“The key thing of it being agentic system of it being an agentic system is that it is autonomous.”

John Roese argues that genuine agentic AI requires an autonomous system capable of reasoning and executing work independently, rather than merely using a chatbot as a tool.

“The key thing of it being agentic system of it being an agentic system is that it is autonomous.”

Reliability is the most critical factor for AI agents, as those that succeed only 60% of the time are essentially useless for real-world production use cases.

“what we observed and and the feedback we've received as well from customers is that on average those agents work maybe 60% of the time. And to be honest, an agent that is reliable 60% is 0% useful, right? Especially if you want to productionize this, you need to really have reliability”

An agent is a large language model with access to tools that allows it to decide on its own path, distinguishing it from a static workflow.

“an agent quite simply is an LLM, a large language model with access to tools. and and and I mean tools in the function calling sense, meaning this thing can Google something if it wants to. This thing can read a file from my desktop if it wants to while it's talking to me.”

Agentic AI is an effective way to bridge the gap between lab development and production by combining reasoning with specialized task execution.

“casting what seems like generative AI problems into agentic problems has allowed us to bridge the gap between the lab and production”

Agentic AI is a major shift because it enables systems to use tools and data autonomously to solve problems rather than just acting as a predefined process flow.

“Agentic AI is not like a process flow diagram. It's giving it a set of tools and letting it go figure out what to do and actually go do it.”