FINALLY! Conversational Commerce is “this close.”

Still, the thought that I could have my own personal virtual assistant that knows past purchases, preferences, payment vehicles and is empowered by me to book travel, order dinner, or unsubscribe from a streaming service invokes a commonly shared memory. It’s the recurring schtick in the “Peanuts” comic strip. Charlie Brown lines up to kick the football, Lucy pulls it away. Charlie Brown yells “AAUGH”.

When it comes to Agentic AI, I’m like Charlie Brown – an optimist thinking that things will be different this time.

The Past is Prologue

For forty-plus years I’ve sought to discover, understand, and encourage adoption of technologies that improve conversations, especially between and among customers and the brands they want to do business with. In the 1980s, toll-free numbers were a big thing, giving rise to the need for “call centers” capable of receiving thousands of calls simultaneously. That gave rise to interactive voice response (IVR) systems to discover the purpose of the call, and automated call directors (ACDs) to route the calls to the “agents standing by.” Back in those days, all agents were live.

As time passed, we’ve seen constant improvement. Yet solutions have consistently fallen short for those of us who expected that, by now, each of us would have access to our own, virtual personal agent (VPA); one that understands what we’re trying to accomplish and then carries out tasks on our behalf. A succession of “new” technologies came along to fulfill those goals. It took on biblical proportions. Natural language processing begat speech recognition, which begat conversational AI, and then generative AI.

Each generation improved on the performance of its predecessors, but they also exposed gaps that required “new new” technologies to tackle. In the case of carrying out conversational commerce, VPA might act autonomously to log onto a business’s web site and take the necessary steps to complete a transaction, or an enterprise’s virtual agent might be called upon to build its own API into a product database to learn of inventory status. These are the sorts of tasks that Agentic AI, and its anthropomorphized offspring, AI Agents, are prepared to perform.

AI Agents Expand Capabilities of LLMs and GenAI

Cobus Greyling, Evangelist at Kore.ai, has called out the features and functions of AI Agents here.

My bulleted shorthand is as follows:

1. They can act autonomously, powered by one or more language models, especially large action models (LAMs), which enable the AI Agent to understand and address complex tasks.
2. They break down problems into sequential steps/sub-tasks, handling each individually.
3. Through iterative cycles of thought, action, and observation, an AI Agent adapts its responses based on feedback.
4. They use tools for interacting with systems like APIs or web searches, enabling them to handle diverse tasks and execute intricate workflows.
5. Each tool has a description in natural language, the AI Agent then matches the sub-task at hand with the tool description to know which tool to match with which sub-task.
6. Tools can include functionality like web search APIs, data retrieval APIs, code execution environments, browser automation tools, natural language processing (NLP) APIs, file management systems, machine vision, etc.

The key take-away is that AI Agents understand instructions in plain English and then carry out all the sub-tasks required to get the job done.

Twenty Years On: Fulfilling on the Original Vision for Siri

Today’s AI agents are the product of an evolutionary process of their own. First came personal assistants, like Siri, Alexa or Google Assistant. When first introduced, I expected the app to act as a mediator on my behalf as a consumer as I did my daily searches. I envisioned occasionally turning to it for shopping and other mobile commerce. In spite of significant advancements in LLMs and GenAI, those expectations are largely unfilled.

The parallels to the development and introduction of what is now Apple’s Siri are uncanny. As I mention in this post from 2010, Dag Kitlaus, Adam Cheyer, and Tom Gruber founded Siri as an independent company in 2007 and were able to leverage research conducted under the auspices of SRI, under the project name CALO (“Cognitive Assistant that Learns and Organizes”) to create the first Siri app, which debuted in the app store in 2010. As noted in this post, just one year later, Siri was more deeply integrated into Apple’s ecosystem offering a list of pre-packaged solutions for selected activities, topped by “Restaurants,” “Movies”, “Events”, and “Local Businesses”.

As I explained back then, “Siri users benefit from a voluminous amount of pre-preprocessing and organization of information that has been carried out ‘in the cloud’ on their behalf.” Communications channels and APIs were established among the likes of OpenTable, Flixster, Taxi Magic and other service providers. Adding new domains and service providers involved time-consuming, laborious and complex processes.

Things will be Different This Time, Right?

Agentic AI could eliminate the time it takes for Siri and its cohort of AI Agents to establish relationships with brands and their sales or customer care systems. But a lot of loose pieces need to fall into place. AI agents are just that: “agents”, which is also a legal term for people authorized to act on behalf of a person. Terms and conditions need to be established to define just what they are empowered to do. There will also be significant issues surrounding the security of personal data (such as preferences) provided to those agents, and quite a lot of details to be worked out surrounding the terms and conditions under which a brand or enterprise’s legal department will recognize instructions given by a live person’s digital representation.

In other words, “AAUGH”.

CX and contact center planners are well-advised to include “agent-to-agent” communications while defining the personnel, policy, technology and procedures they employ in their customer care contact centers.

Categories: Conversational Intelligence, Intelligent Assistants, Articles