You’ve heard the buzz about AI agents, but what is an AI agent, really? Think of them as digital assistants with superpowers, ready to revolutionize how we work and live. In this article, I go through a brief definition and present overview with a view on how AI agents will impact us over the next few years. I highly recommend checking out the links and video in this article if you are interested to learn more.
1. What is an AI Agent?
An AI agent is a program designed to perform tasks intelligently and autonomously, often mimicking human decision-making. Think of it like a person who has been hired to do a specific job. This could be a small job or a big job. An AI agent is armed with a directive and a specific set of tools to do the job. It uses algorithms to analyze data and make decisions and can often learn and improve over time. For me, I think of the robot that was designed in an episode of Rick and Morty that was given a sole purpose of passing butter.
Robot: “What is my purpose?”
Rick: “You Pass Butter”
”Robot:[Looks at its hands] “Oh my God.”
Rick: “Yeah Welcome to the Club, Pal.”
While humorous, the ‘Pass the Butter Robot’ highlights the simplicity and purpose-driven nature of early AI agents. Modern AI agents, however, are far more versatile and capable of handling complex tasks, thanks to advancements in machine learning and data processing. AI agents, in essence, are programs that act autonomously and intelligently to solve problems or perform tasks.
2. How AI Agents Work
Agents work from Inputs, then process the information and provide some kind of an output. An input can be from user commands, from a previous step in a workflow, or a data from a sensor for example. Processing will be determined by the tools it has available, and it can also be AI techniques like machine learning and natural language processing to analyze inputs. The output will be an action or response.
For example, an AI customer service agent takes an input such as a user query (‘Where is my order?’), processes it using natural language processing, and generates an output by retrieving and displaying the order status.
You can see this is a very general description of a process that involves an agent. Google recently came out with an Intro to AI agents YouTube video that says “no one seems to agree on what an agent is.” It goes on to focus on AI tools having “Agentic” like properties which is acting autonomous essentially.
There is another good resource referenced in the above video written by Andrew Ng. Andrew Ng is the co-founder of Google Brain and Coursera, and a former Chief Scientist at Baidu, Andrew has dedicated his career to making AI accessible to everyone and is one of the most influential figures in the field of artificial intelligence and machine learning. His blog is a description of what makes agentic behavior broken down into 5 parts and is worth the read if you’re interesting in understanding the basics further. I won’t pretend to break it down better than he can, I’m still learning. You can read his 5 part article here.
3. Real-World Examples of AI Agents
Voice Assistants: Alexa, Siri, Google Assistant.
Customer Support Bots: Chatbots on websites that help with questions or troubleshoot issues.
Autonomous Vehicles: Self-driving cars making decisions on the road.
Personalized AI Agents: Tools like Notion AI or GitHub Copilot that enhance productivity.
Creative Tools: GitHub Copilot, Jasper AI.
What others do you know? I think that well-crafted prompts are agents as well. We’ll explore those in more detail as well, in the mean time feel free to head over to this subreddit to explore the latest and top posts: ChatGPTPromptGenius.
4. Why AI Agents Matter for the Future
Fortunately, or unfortunately, AI armed with the ability to iterate itself, learn and use tools will easily replace a good portion of the workforce that’s designed to work behind a computer. It’s one of the reasons I feel the need to learn AI and feel existential dread every day as a solution architect of a SaaS company. As of today, several companies have stated that for 2025 they will not be hiring as many engineers or data analysts because of their expectation to be able to leverage AI, including AI agents.
The landscape of working is changing. I don’t have the philosophical, knowledge and perspective at the moment to properly describe its impact, but it’s clear that an AI Agents will replace many jobs that are done by humans. I’m hoping that, like technological advances in the past, humans find a way to leverage and grow with the technology as opposed to be replaced 1 for 1. Much like the Industrial Revolution shifted workers from repetitive labor to specialized roles, AI agents are poised to redefine how we work. They automate mundane tasks, creating opportunities for humans to focus on innovation, creativity, and strategy
With regards to the Industrial Revolution, at first, these changes caused disruption, as people worried that machines would entirely replace human labor. However, instead of rendering workers obsolete, the new technology created opportunities for humans to specialize and take on higher-value roles. For instance:
- In Textiles: Skilled artisans transitioned from hand-weaving to operating and maintaining machines, which boosted productivity and lowered costs.
- In Agriculture: Mechanized tools reduced the need for manual farm labor, enabling workers to shift to urban centers where industries like manufacturing and services flourished.
- In Transportation: Steam-powered trains and ships not only created jobs for engineers, machinists, and conductors but also expanded commerce, creating demand for workers in other industries.
Just like the machines of the Industrial Revolution, AI agents are tools designed to improve efficiency and handle repetitive or mundane tasks. While they may replace certain roles, they also open the door to new possibilities:
- Instead of spending hours extracting and organizing data manually, workers can focus on analyzing and interpreting it.
- AI can handle routine customer inquiries, allowing customer service representatives to address complex problems and build better relationships.
- In creative fields, AI can assist with ideation or technical tasks, giving humans more room for artistic expression and innovation.
We see currently several examples of AI-driven roles emerging as new job opportunities. For example, AI trainers, prompt engineers and ethical AI consultants are roles that exist today but did not exist a just a few years ago.
I found this video both fascinating and, frankly, intimidating. Microsoft’s CEO, Satya Nadella, explains that apps as we know them are dead, and AI agents are poised to take over. In the video below, Matthew Berman breaks this concept down in a way I could understand. My takeaway from the video is that there will be many AI agents, smart companies are investing heavily in them, and they are set to redefine the architecture of many software companies. There are far more insights in the video than I can summarize here, and it’s definitely worth a watch. Please comment if you found anything particularly impactful—I’d love to discuss it further.
That said, I do feel uneasy about the current limitations of AI, especially its flimsiness compared to the structured, hardcoded reliability of traditional databases. I’ve seen too many errors AI makes in seemingly simple tasks—like when AI struggled to correctly count the number of “r”s in the word strawberry. However, the pace of AI improvement is undeniably impressive, and I don’t want to underestimate its potential, no matter how cautious I may feel.
I was joking with my wife today—somewhat despairingly—and said that, in the near future, the new catch phrase is going to be “Support Humans!” much like how “Buy Local!” has been in recent years. The new farm-to-table movement might just be some old geezer like me (hah) cranking out an excel spreadsheet.
In future articles, I’ll explore creating simple AI Agents with you. In the meantime – stay ahead of the curve and check out our guide on ‘Top 9 Free AI Courses to Master Artificial Intelligence in 2025.‘ If you have an idea or request for an AI agent you created, use or would like to see, please let me know in the comments.