What future do we envision for artificial intelligence? The answer isn’t simple, since we’re talking about a tool with incredible potential, and we’ve only just begun to explore its capabilities. But, clearly, we’re thinking about something that goes far beyond simply providing answers (however good and useful they may be) in a chat.
For most people we ask this question, the future of this technology lies in solving our problems and performing tasks autonomously and automatically. That future is already here. And it’s known as agentic AI. Simply put: it’s artificial intelligence that goes from talking to doing.
Agent AI? What is that?
Perhaps the name isn’t the most appealing aspect of this technology. But in terms of utility, it stands out above other resources. The main idea is that this artificial intelligence model, unlike generative AI, has the ability to act to achieve specific goals that we assign it: it makes decisions and takes action.
That can be frightening, to a certain extent. And it might remind us of scenes from some of the most famous science fiction films in history. But, as we mentioned before, we humans assign the objective: we define the specific task it must complete. And the agentic AI will carry it out independently and autonomously (within the limits we set).
In short, it will plan, execute, and learn to improve. The possibilities, therefore, are vast and promising. If generative intelligence could suggest a musical performance in our city, agentic AI can suggest the performance, book the tickets and a table for dinner afterward, arrange a taxi for the round trip, and so on.
Now, let’s apply this example to other tasks. For instance, an automated agent that buys shares in investment funds based on our criteria. Or one that runs a digital marketing campaign based on changes in key performance indicators. Or one that blocks cyberattacks on our system by anticipating the methods used by hackers.
An invaluable resource for businesses
Clearly, this is a giant leap for businesses. It means streamlining tasks and reducing the margin of error; but also saving costs and freeing up human talent for higher-value tasks (including setting the objectives of agentic AI and monitoring the results). If generative AI was already a significant contribution, this is a breakthrough of enormous magnitude.
Let’s consider, for example, a company that wants to optimize its purchases of raw materials and services. The AI agent can be trained to contact different suppliers, request rates, compare them, ask for specific quotes (including shipping, warranty period, etc.), make decisions, and execute the process. It can even verify deliveries and file claims or make changes.
Let’s look at a specific example in a particular sector. Online casino games are among the most common digital products on the internet; however, by their very nature, they require proven and fail-safe functionality. Agentic AI can subject them to tests of thousands of games, verifying that all parameters (graphics, RTP, purchase of special features, etc.) function perfectly and suggesting improvements before they are released to the market.
The great ally of Canadian entrepreneurs?
In Canada, some institutions are already applying this technology to specific tasks and activities. The Ontario Tech University’s learning agent generates assistants for students that personalize study materials. Another example is that of a well-known web design company, which creates and implements user-suggested designs while resolving potential problems.
Of course, agentic AI can be crucial for new projects and for boosting entrepreneurs. Beginnings are always difficult and require resources that not everyone has. A technical tool that automates vital tasks for project development democratizes access to the market and customers. Innovative ideas now have a new and valuable ally to become reality.
