The Evolution of AI: From Chatbots to 'Agentic AI'

The Evolution of AI: From Chatbots to "Agentic AI"

If 2023 was the year I learned to talk to AI, and 2024 was the year I learned to build with it, 2025 is shaping up to be the year AI starts doing the work for me.

For the past two years, the world has been captivated by LLMs (Large Language Models). These "chatbots" act as incredibly knowledgeable consultants—they can write a marketing plan or debug code. But there is a catch: once the chatbot gives you the answer, you still have to execute the task.

That dynamic is about to change. I am witnessing the rise of Agentic AI.

What is Agentic AI?

In simple terms, "Agentic AI" refers to AI systems that possess agency. Unlike a traditional chatbot, which waits passively for a prompt and responds with text, an AI Agent is designed to pursue goals, reason through complex problems, and—crucially—take action to achieve a result.

• The Chatbot (The Consultant): You ask, "How do I optimize my server costs?" It gives you a checklist of steps. You have to log in and do them.

• The Agent (The Employee): You say, "Optimize my server costs." The agent logs in, analyzes the usage logs, identifies idle resources, creates a backup, shuts down the waste, and sends you a report of the money saved.

It isn't just generating text; it's generating outcomes.

The Impact on FinTech and Development

In Financial Technology (FinTech)

I am moving beyond simple "robo-advisors." Agentic AI in FinTech means autonomous remediation. Imagine an agent that monitors your business cash flow 24/7. When a vendor invoice arrives, the agent acts: it verifies the invoice, checks liquidity, schedules the payment to maximize interest, and logs the transaction.

In Web Development

For developers, the "Copilot" era is evolving. I am starting to build agents that act as autonomous engineers. Instead of just completing a line of code, an agent can be assigned a Jira ticket. It can read the requirements, write the code, write the test cases, and deploy the fix to a staging environment.

The Future is Autonomous

The transition to Agentic AI is not just a technical upgrade; it's a fundamental shift in how I work. It allows me to offload the "doing" to machines so I can focus on the "deciding."

Best,

Gerasimos Makris Founder of g-makris.com AI Web Developer | Double Master's in MBA & FinTech and Blockchain

Tech Glossary & Concepts

• Agentic AI: An AI system capable of autonomous action to achieve a specific goal with minimal human intervention.
• LLM (Large Language Model): A deep learning algorithm that can recognize, summarize, translate, predict, and generate content using very large datasets.
• Agency: The capacity of an actor to act in a given environment. In AI, it means the ability to take initiative.
• Remediation: The process of correcting or fixing something that is wrong or damaged. In tech, it usually means fixing security vulnerabilities or compliance issues.