AI Agents: The Next Big Thing After Chatbots
Let's get you started.
Hi, happy Friday! :)
Remember when OpenAI introduced curated and custom GPTs within ChatGPT to make specific and possibly repetitive tasks easier?
That was just the beginning. Now, AI Agents are stepping up as the next big thing in AI, offering powerful automation capabilities.
An Amazon Web Services article defines AI agent as:
A software program that can interact with its environment, collect data, and use the data to perform self-determined tasks to meet predetermined goals. Humans set goals, but an AI agent independently chooses the best actions it needs to perform to achieve those goals.
AI agents behave rationally. They make sensible, autonomous decisions based on their observations and data, resulting in optimum performance and results. This is the primary difference between ChatGPT/custom GPTs and AI agents.
An AI agent detects its surroundings via physical or software interfaces, and a large language model (LLM) is integrated by it together with improved prompting mechanisms and external functions.
Functions of AI Agents
Recognise Dynamic Changes: Excel at identifying changes in dynamic environments, such as customer behaviour on an e-commerce site or data shifts in financial markets.
Adaptive Actions: React to changes in real-time. For eg, they can adjust prices on an e-commerce site based on consumer behaviour or market trends.
Interpretation and Reasoning: Beyond data collection, they make decisions by interpreting complex datasets and extracting valuable insights.
Problem-Solving: Process vast amounts of data swiftly, recognise patterns accurately, and adapt through continuous learning, proving invaluable in fields like financial trading and supply chain optimisation.
Inference and Learning: Analyse past and present data to predict future outcomes, improving performance through learning from each interaction.
Action and Outcome Analysis: Plan and execute actions to achieve desired goals, considering various scenarios and their implications, especially in strategic planning and decision-making.
AI Agents You Should Try
Ready to dive in? Let’s get you started :)
There are a few ways to get started with AI agents:
Build Your Own: Using Python and a large language model (e.g., OpenAI API, Anthropic, Gemini), you can code your own AI agent. For a detailed guide, check out Hasan Aboul Hasan’s article, "How To Create Autonomous AI Agents From Scratch!"
Use a Third-Party App: AI Agent is a straightforward app that allows you to customise and create your own agents to automate tasks. You can set its name, language model, and goals. It generates an initial task list which you can further edit and refine.
Explore Project Astra: Announced at Google’s recent I/O event, Project Astra is a real-time AI agent designed to assist consumers in navigating the world and completing tasks. It can recognize objects, explain code, locate you in the city, and even find misplaced items. Check out the impressive demo video here.
And that’s it! I hope you try AI agents out while they’re still early.
All of Big Tech is tapping on this opportunity, so don’t sit on it. Try it, see how you like it, and customise it to make it work for you :)
Have questions? Let’s chat in the comments!