Customer Experience Trend: The Rise of AI Agents

Home Customer Experience Trend: The Rise of AI Agents

Never in my life have I experienced a technological acceleration as exponential as that of AI, and more specifically of Generative AI (GenAI), over these past 2 years. If we can believe Sam Altman, we are just a few thousand days removed from Artificial General Intelligence (AGI), which he defines as intelligence that is comparable to a median human co-worker, capable of learning and excelling in diverse fields.

This is how Altman sees GenAI evolving:

  1. Stage Level 1: Chatbots, AI with conversational language
  2. Stage Level 2: Reasoners, human-level problem solving
  3. Stage Level 3: Agents, systems that can take actions
  4. Stage Level 4: Innovators, AI that can aid in invention
  5. Stage Level 5: Organizations: AI that can do the work of an organization

We are currently firmly rooted in level 1, though we are moving towards level 2 with reasoning models (which are good at science, coding, and mathematics) like OpenAI’s o1 which is learning to perform complex reasoning. It always strikes me how most of us seem to have become pretty used to the skills of these conversational chatbots, which just goes to show how fast they are evolving and how quickly they became normal to us.

Indistinguishable from humans

Not so long ago, Salesforce CEO Marc Benioff claimed that AI has advanced to a point where it is indistinguishable from humans in customer service roles. That will of course have (and sometimes already has) a huge effect on customer service. Over the past year, for instance, payment provider fintech Klarna has been very vocal about its use of GenAI in customer environments. After one month of using its OpenAI-powered customer support agent, the latter was handling two thirds of customer chats, the equivalent of 700 human agents, with the same accuracy but a lot faster.

But conversational chatbots are only one part of the equation. If AI keeps evolving as fast as it is now, AI agents might not take long to mature. Agentic AI will not only have a huge impact on our customer interactions, it will also force us to revisit our websites, our marketing or our interactions with employees.

For those who aren’t so familiar with artificial intelligence (AI) agents: they are a system or program that is capable of autonomously performing tasks on behalf of a user or another system. An AI chatbot answers questions, but an AI agent acts on your behalf.

If you think that these agent systems will only manifest in a far of future, realize that over the past weeks and months, several of the big AI companies made announcements in the matter:

  • Anthropic launched a model that can control your screen to do tasks in your apps
  • Asana launched a no-code tool for designing AI agents
  • Microsoft has added a layer of ‘autonomous agents’ to some of its enterprise tools
  • Google is planning to launch an AI agentic tool codenamed Project Jarvis (which it mistakenly briefly released but then retracted)
  • Salesforce announced the general availability of its AI agent development platform called Agentforce.
  • Apple filed several patents around AI agents
  • Jensen Huang, CEO of Nvidia, envisions his company having 100 million AI agents.

The ultimate friction removing tool

Though this is just the beginning and agentic AI still needs to mature, when it comes to CX, there are 2 ways that companies need to start thinking about it:

  1. What will agents mean for their customer relations and interactions
  2. What will they mean for their marketing

When it comes to CX, I believe that AI agents will be the ultimate friction removing tool, taking away a huge number of hurdles.

Let’s say you want to organize a road trip in Australia. Today, you must do research about interesting places and people to visit (possible in ChatGPT and consorts), then go to a website or app to book a flight, then to another one to book your different hotels, then to another one to book a car, a restaurant, a museum… Soon, you will ask your agent and they will manage that entire journey for you in a highly personalized manner, requiring zero-click interaction from you. In fact, agents may become proactive at one point, calling your hotel that you will be later because you are still 150 km away, while you were scheduled to arrive in 15 minutes.

Agent-to-Agent Interaction

These agents may trigger an intense paradigm shift when it comes to your customer interactions. At first, they will interact directly with customers over your website, social media channels or call centres. But we might get to a point where customer agents interact with organizational agents.

On a deeper level, this also means that companies should redesign their websites, apps, and other digital platforms to be accessible and usable by AI agents, if those will become primary “visitors” to their websites. They might even need to develop dual interfaces: one optimized for human customers and another for AI agents that prioritizes structured data and programmatic interactions. This may, according to my friend Jeremiah Owyang, result in a new type of economy, with new social norms. Where AI agents will interact autonomously, (trading, buying and selling) and evolve (learning, governing, and reproducing) with minimal human intervention.

This is truly an exciting area. The advantages of advanced chatbots and, later, AI agents seem (and will be) tremendous for companies when it comes to employee augmentation (especially in tedious and repetitive tasks), 24/7 accessibility, productivity, speed, personalization and friction removal. It’s a tremendous opportunity, but also a challenge. Those companies that view chatbots and agents purely as efficiency and productivity enhancers – and that will happen, mark my words – will lose out on customers. What they should be doing is thinking one thought alone: “How will this make my customers happier?”. All the rest is noise.

Play video