Agentic AI – 1
Note: I want to thank Gautam Mehra, co-founder and CEO of ProfitWheel, for introducing me to the world of Agentic AI.
Volodymyr Zhukov writes: “Agentic AI marks a fundamental shift in the very nature of artificial intelligence. It describes a class of AI systems specifically designed to understand complex workflows and pursue intricate goals autonomously, with little to no human intervention. In essence, agentic AI functions more like a human employee, grasping the complex context and instructions provided in natural language, embarking on goal-setting, reasoning through subtasks, and adapting decisions and actions based on changing conditions… Agentic AI can automate and optimize operations, adding significant value to business processes. These enhancements go beyond mere automation, offering advanced problem-solving capability and strategic planning to tackle complex tasks.” He lists the features that set it apart:
- Autonomy: Unlike traditional AI systems, agentic AI is built to take initiative, performing directed actions independently, without constant human supervision.
- Reasoning: Agentic AI possesses an advanced degree of decision-making, allowing it to make contextual judgments, weigh trade-offs, and set strategic actions.
- Adaptable planning: In dynamic and changeable conditions, agentic AI demonstrates flexibility and responsiveness, adjusting its goals and plans based on the prevailing circumstances.
- Language understanding: With an advanced ability to comprehend and interpret natural language, agentic AI can meticulously follow complex instructions, enhancing its capability to tackle sophisticated operations.
- Workflow optimization: Agentic AI exhibits an uncanny skill to move fluidly between subtasks and applications, executing processes with optimum efficiency while ensuring the end goal is achieved.
Ken Rheingans writes: “[AI Agents] are teams of AI-powered virtual assistants that collaborate to solve complex problems. Unlike traditional AI systems that generate a single output based on a given prompt, AI Agents work together, sharing goals and making collective decisions to tackle tasks more effectively. This collaborative approach allows for more sophisticated interactions and decision-making processes, ultimately leading to improved efficiency and significantly more accurate results across various applications. Agentic Workflows advance the concept of AI Agents by incorporating iterative refinement and feedback mechanisms. In an Agentic Workflow, AI Agents can generate drafts, receive guidance to improve those drafts, and iterate on their output. This process can lead to more accurate and refined results, as the AI Agents receive specific feedback and have the opportunity to adjust within the parameters of the assigned task. This improved accuracy is the amazing part where good LLM outputs become superhuman great outputs.”
Margo Poda writes: “Enterprise-wide use cases require more than just well-thought-out responses — enterprises need AI agents that can reliably manage complex goals and workflows. This demand has driven the emergence of agentic capabilities like autonomous goal-setting, reasoning, decision-making, robust language understanding, and the ability to connect with enterprise systems using plugins. These agentic capabilities unlock a new generation of enterprise AI solutions — including AI copilots. These tools are being designed to operate without constant human oversight across varied domains. In this way, agentic systems interpret instructions more accurately, set subgoals to accomplish multi-step tasks, and make adaptive choices adjusting to real-time developments, enabling reliable automation of convoluted business objectives.”
Economist: “There is a way…to make large language models (LLMs) perform such complex jobs: make them work together. Researchers are experimenting with teams of LLMs—known as multi-agent systems (MAS)—that can assign each other tasks, build on each other’s work or deliberate over a problem in order to find a solution that any one, on its own, would have been unable to find. And all without the need for a human to direct them at every step. Teams also demonstrate the kinds of reasoning and mathematical skills that are usually beyond stand-alone AI models. And they could be less prone to generating inaccurate or false information.”
