On the heels of all the Generative AI hype, the talk of the digital town these days is all about AI Agents. While most of us are still trying to comprehend how ChatGPT and the other LLMs work, and if the posts we see on social media are real (they are not) or something whipped up by an AI model (they are), the rest of the technology world is busy thinking about the next ‘big thing’. And they have settled on the subtle term ‘Agents’. Nothing mysterious about that, right?

Real Life Agents

An AI Agent is not a James-Bond-like character crafted from the digital ethos to work in the shadows of cyber security, moving undetected from propaganda crime scenes to the latest island built by the mega-villains. Although this would be a much better use of the term AI Agent. Instead, AI Agents of our near future will not be fighting evil in the name of the Majesty, rather they will be doing the real dirty work. And by that, I mean buying this week’s groceries, or complaining to the city about your erroneous water bill, or maybe ordering a pizza because that event where your kids were singing and dancing was admittedly adorable, it also ran very very very long and threw off your entire dinner plan. Oh the glamor of being an agent…in real life.

…when the AI agents come they can do all of that dialing, talking stuff for you

In other words, the AI Agent will do things for you automatically, like when you search for the number of your mechanic and then have to do that human of all things, dial it on your phone and talk to a real person. No more! Or at least when the AI agents come they can do all of that dialing, talking stuff for you, which will free up your time for some other tasks like folding laundry. Robots and their AI agents will never figure out how to fold laundry (really this is true).

Agent vs. Agent

Of course there is just one other problem. What if your pizza pie guy, or your Molly Mechanic has their own AI Agent?

Part of the current hype around generative AI is that companies can improve the productivity of their workers. According to a recent McKinsey & Company study, 71% of companies are using Generative AI in at least one function. The most popular application is in marketing and sales, with 42% reporting its use in this function. On the low end is supply chain management and marketing, with 7% and 5% respectively. This may be a result of the use of traditional AI and decades of its application in those areas where generative AI has little value to add.

While the use of Gen AI in marketing is no surprise with (better?) highly targeted micro campaigns, it is a precursor to how it will influence AI Agents acting on behalf of the customer. The traditional marketing funnel that starts the customer journey from a need towards a product selection, changes with AI agents performing many of the interim steps, like looking at alternative choices, where the customer does not need to be involved.

…developing a corporate AI Agent at enterprise scale, will be more of a task of years rather than months

AI Agents are Coming for You

AI Agents will evolve quickly on the customer side of the equation with the support of GPT platforms like ChatGPT (GPT-4o), Anthropic (Claude Opus, Claude Sonnet), Microsoft (Copilot). OpenAI’s Operator is already available as a preview product and shows impressive capabilities, such as this demo finding a recipe and ordering the needed groceries.

For many companies with online services, developing a corporate AI Agent at enterprise scale that can respond to customers’ AI Agents, will be more of a task of years rather than months. This technology is moving very fast and the question is: Will companies be ready for AI agents and are they developing those strategies now?

The agents are coming.