Artificial intelligence first came to light in the mid-1950s as a promising new academic discipline that aimed to develop intelligent machines capable of handling human-like tasks like natural language and playing chess. AI became one of the most exciting areas in computer sciences in the 1960s, ’70s, and ’80s, but after years of unfulfilled promises and hype, a so called AI winter of reduced interest and funding that nearly killed the field set in everywhere.
AI was successfully reborn in the 1990s with a totally different data centric paradigm based on analyzing large amounts of data with sophisticated algorithms and powerful computers. We were wowed when in 1997 Deep Blue won a celebrated chess match against then reigning champion Gary Kasparov, — one of the earliest and most concrete grand challenges of AI. The 2010s saw increasingly powerful deep learning AI systems surpass human levels of performance in a number of tasks like image and speech recognition, skin and breast cancer detection, and winning at championship-level Go.
Then came Generative AI. In the early 2020s, the impressive ability of GenAI large language models (LLMs) and chatbots to understand our prompts and generate cogent, articulate sentences has given us the illusion that we’re dealing with an educated, intelligent human, rather than with a highly sophisticated technology that’s been trained with huge amounts of human language.
AI has now emerged as one of, if not the key defining technology of the 21st century.
Next in AI’s progression is the evolution from knowledge to action based on Agentic AI, — “a class of artificial intelligence that focuses on autonomous systems that can make decisions and perform tasks without human intervention.”
A July, 2024 McKinsey report nicely explained the difference between generative and agentic AI. “We are beginning an evolution from knowledge-based, gen-AI-powered tools — say, chatbots that answer questions and generate content — to gen AI-enabled agents that use foundation models to execute complex, multistep workflows across a digital world.”
Similarly, in “The 2025 Guide to AI Agents,” IBM wrote: “An artificial intelligence (AI) agent refers to a system or program that is capable of autonomously performing tasks on behalf of a user or another system by designing its workflow and utilizing available tools. AI agents can encompass a wide range of functionalities beyond natural language processing including decision-making, problem-solving, interacting with external environments and executing actions.”
And a recently published book, Agentic Artificial Intelligence: Harnessing AI Agents to Reinvent Business, Work and Life, included a very good overview of the value of Agentic AI in its Introduction, particularly its explanation of the difference between Generative and Agentic AI based on three concrete use cases.
But, “cracks in the agent armor have begun to appear,” noted Babson College professor Tom Davenport in a July 7 article on “The Roles of Humans in an Agent-Driven World,” a co-author of the Agentic AI book. “Everybody in the AI world is excited about agents, which of course can perform digital tasks of various types rather than just inform curious humans. However, astute observers are a bit less excited about agents now than they were a few weeks ago.” For example:
- Gartner predicts that 40% of agentic projects will be canceled by 2027. “Surprising that a technology in its infancy is already entering Gartner’s trough of disillusionment.”
- Anthropic ran an experiment to see if an AI agent could run a small automated shop. The experiment did not go well; the AI agent “made too many mistakes to run the shop successfully.”
- Carnegie Mellon researchers tried out ten different agents to run a software business. “The agents made lots of mistakes; the most effective agent only completed 24% of its assigned tasks successfully.”
While acknowledging that AI agents will get better over time, and that human workers also make mistakes, Davenport added that “these predictions and results suggest a specific set of roles for humans in a world of lots of AI agents. Are they roles that we want to play? I can’t speak for all humans, but I’m not sure I would want these jobs. Here are some I envision:”
- Human agents — “The inability of agents to perform all roles suggests that there will be ‘human agent’ roles to do certain steps in a process that an AI agent can’t (yet) perform well.” For example, in an Amazon warehouse, humans might perform tasks like unloading trucks and packing orders of unusual sizes or shapes.
- Agent auditors — “If AI agents make a lot of mistakes, they will need humans to review and audit their performance. They’ll also probably have to reverse or fix the mistakes that agents have made.”
- Incremental agent improvers — “A well-designed work environment involving AI agents would probably require that humans not only monitor and correct their mistakes, but also make suggestions to the agents (or their human creators) for how to improve their performance. In order to do this well, the human worker would need to be technically astute and know something about how the agents work.”
- Agent architects — “One of the better roles in an agent-centered world is designing what they do in an overall system or process. I don’t think it will be easy, given that a broad process may involve lots of different agents, each doing a different task. And given the current agent accuracy level, it will involve both human and AI actors.”
- Agentic decision-makers — “The most important role in an agent-centered world is deciding whether to use agents or not for a particular purpose. Again, this could be a tough job. Agents are likely to put some humans out of work, and being a hatchet person is never fun.”
- Agentic psychologists — “A somewhat tongue-in-cheek role I might suggest is that of the agentic psychologist, who would counsel unhappy humans about their daily work with AI agents. AI can usually work faster than we humans can, and doesn’t need to take breaks. The digital tasks that humans still perform are also likely to be measured. The speed and monitoring of the AI/human collaboration will probably put stress on the humans assigned to partner with them.”
“I would stress that it’s early days for agentic AI, and we don’t know how things will eventually turn out in terms of the capabilities of agents and the ways humans will need to adapt,” wrote Davenport in conclusion. “But perhaps we should think a bit about the implications for humans before we charge off too fast in an agent-focused direction.”
How are AI agents likely to evolve into the future? Are they likely to rely on human collaboration, become increasingly autonomous, or both?
In a recent blog, Agents in the AI-First Company, entrepreneur, VC, and author Evangelos Simoudis introduced a five-level AI Agency spectrum. I’ve known Simoudis since the 1990s when he spent a few years in IBM as VP of business intelligence. Over the past decade he’s closely followed the evolution of the auto industry and wrote a 2017 book, The Big Data Opportunity in Our Driverless Future. So, not surprisingly, his five-level AI Agency spectrum is modeled after the five-levels of driving automation defined by the National Highway Traffic Safety Automation.
The different agent levels and their associated human roles are:
- Level 1: Basic Automation. “The agent follows predefined deterministic rules to perform tasks, but it does not learn from its environment or adapt its behavior; — the human operator invokes the agent for a specific tasks and interprets the resulting output.”
- Level 2: Conversational Assistant. “The agent can understand input and generate relevant output for a single session, but it lacks persistent memory and the ability to learn independently; — the human guides the agent through a continuous dialogue of prompts and questions.”
- Level 3: Contextual Adaptive Agent. The agent maintains memory during a task and adjusts its actions in real-time, but learning is localized and limited to the immediate context; — the human supervises the agent, sets the overall task goal, and intervenes if necessary.
- Level 4: Autonomous Learning Agent. The agent is capable of learning from its actions and applying that knowledge to future tasks, and it can evolved its strategy over time without human intervention; — the human architect defines the objectives and learning setup, and oversees the agent’s evolution.
- Level 5: Collaborative Multi-Agent System. Multiple Level 4 agents work together in a shared environment and the system exhibits collective intelligence to solve complex problems; — the human strategist sets the high-level mission for the entire system and monitors the collective outcome.
“Enterprises are actively experimenting with a variety of internally developed and third-party AI agents, wrote Simoudis. “Agents already deployed in production, from a Waymo vehicle to OpenAI’s Operator, show how far we have come, but also how far we still have to go. For the corporation, agentification is not a one-time decision; it is a maturity journey.”