Comments about "Intelligent agent" in Wikipedia

This document contains comments about the article Intelligent agent in Wikipedia
In the last paragraph I explain my own opinion.

Contents

Reflection

Introduction

The article starts with the following sentence.
In artificial intelligence, an intelligent agent is an entity that perceives its environment, takes actions autonomously to achieve goals, and may improve its performance through machine learning or by acquiring knowledge.
That sounds promising, but the knowledge lies in the details. That is missing.
It is very important how an intelligent achieves it goals. This should be described in detail in a document.
The term autonomously indicates that there exists no way to influence the behavior of the AI agent. That seems dangerous.
AI textbooks[which?] define artificial intelligence as the "study and design of intelligent agents," emphasizing that goal-directed behavior is central to intelligence.
A specialized subset of intelligent agents, agentic AI (also known as an AI agent or simply agent), expands this concept by proactively pursuing goals, making decisions, and taking actions over extended periods.
The four concepts intelligent agents, agentic AI, AI agent and rational agent are confusing concepts.
The confusion starts because they seem to be same, but apperently they are not.

Intelligent agents can range from simple to highly complex.
A basic thermostat or control system is considered an intelligent agent, as is a human being, or any other system that meets the same criteria—such as a firm, a state, or a biome.
A basic thermostat or (classical) control system have nothing to do with artificial intelligence.
A human being also has nothing to do with artificial intelligence.
See also: Reflection 1 - AI agent versus intelligent agent

Intelligent agents operate based on an objective function, which encapsulates their goals. They are designed to create and execute plans that maximize the expected value of this function upon completion. For example, a reinforcement learning agent has a reward function, which allows programmers to shape its desired behavior.
This example creates more problems/questions than it solves.
Similarly, an evolutionary algorithm's behavior is guided by a fitness function.
Intelligent agents in artificial intelligence are closely related to agents in economics, and versions of the intelligent agent paradigm are studied in cognitive science, ethics, and the philosophy of practical reason, as well as in many interdisciplinary socio-cognitive modeling and computer social simulations.
Intelligent agents are often described schematically as abstract functional systems similar to computer programs. To distinguish theoretical models from real-world implementations, abstract descriptions of intelligent agents are called abstract intelligent agents.
Computer programs are only schematically discussed when they are in a design phase.
The use of the concept abstract is confusing.
Intelligent agents are also closely related to software agents—autonomous computer programs that carry out tasks on behalf of users. They are also referred to using a term borrowed from economics: a "rational agent".[1]
Classical computer programs, all the programs in a PC, perform narrowly defined tasks under control of users. The program I'm using allows me to put my thought on paper. The functionality (operation) of the program is clearly described in a description of the program. This functionality has nothing to do with artificial intelligence nor with AI agents. The use

1. Intelligent agents as the foundation of AI

The concept of intelligent agents provides a foundational lens through which to define and understand artificial intelligence. For instance, the influential textbook Artificial Intelligence: A Modern Approach (Russell & Norvig) describes:
Neither the text above or the text below are clear. The question what is artficial intelligence and what is not is difficult. The introduction of an agent does not make the subject easier.
Agent: Anything that perceives its environment (using sensors) and acts upon it (using actuators). E.g., a robot with cameras and wheels, or a software program that reads data and makes recommendations.
To do anything with this information requireas human input and reasoning.
Rational Agent: An agent that strives to achieve the *best possible outcome* based on its knowledge and past experiences. "Best" is defined by a performance measure – a way of evaluating how well the agent is doing.
Same as above. Requires human input and reasoning.
Artificial Intelligence (as a field): The study and creation of these rational agents.
Other researchers and definitions build upon this foundation. Padgham & Winikoff emphasize that intelligent agents should react to changes in their environment in a timely way, proactively pursue goals, and be flexible and robust (able to handle unexpected situations).
Same as above. Requires human input and reasoning.
Some also suggest that ideal agents should be "rational" in the economic sense (making optimal choices) and capable of complex reasoning, like having beliefs, desires, and intentions (BDI model).
All of this requires human input and reasoning.
Kaplan and Haenlein offer a similar definition, focusing on a system's ability to understand external data, learn from that data, and use what is learned to achieve goals through flexible adaptation.
The real effect lies in the detail.
Defining AI in terms of intelligent agents offers several key advantages: Avoids Philosophical Debates: It sidesteps arguments about whether AI is "truly" intelligent or conscious, like those raised by the Turing test or Searle's Chinese Room.
The question of what is the difference between intelligent and articial intelligence requires to be asked and answered. Artificial intelligence is never conscious.
It focuses on behavior and goal achievement, not on replicating human thought.
It requires a clear distinction between mechanical and human based solutions.
Objective Testing: It provides a clear, scientific way to evaluate AI systems.
This requires
Researchers can compare different approaches by measuring how well they maximize a specific "goal function" (or objective function). This allows for direct comparison and combination of techniques.
Interdisciplinary Communication: It creates a common language for AI researchers to collaborate with other fields like mathematical optimization and economics, which also use concepts like "goals" and "rational agents."

2. Objective function

An objective function (or goal function) specifies the goals of an intelligent agent. An agent is deemed more intelligent if it consistently selects actions that yield outcomes better aligned with its objective function. In effect, the objective function serves as a measure of success. The objective function may be:
Simple: For example, in a game of Go, the objective function might assign a value of 1 for a win and 0 for a loss.
Complex: It might require the agent to evaluate and learn from past actions, adapting its behavior based on patterns that have proven effective.
The objective function encapsulates all of the goals the agent is designed to achieve. For rational agents, it also incorporates the trade-offs between potentially conflicting goals. For instance, a self-driving car's objective function might balance factors such as safety, speed, and passenger comfort. Different terms are used to describe this concept, depending on the context. These include: Utility function: Often used in economics and decision theory, representing the desirability of a state.
Objective function: A general term used in optimization.
Loss function: Typically used in machine learning, where the goal is to minimize the loss (error).
Reward Function: Used in reinforcement learning.
Fitness Function: Used in evolutionary systems.
Goals, and therefore the objective function, can be:
It is rather simple to describe all these functions in theory, but very difficult in practice baised on real applications.
Explicitly defined: Programmed directly into the agent.
Induced: Learned or evolved over time.
In reinforcement learning, a "reward function" provides feedback, encouraging desired behaviors and discouraging undesirable ones. The agent learns to maximize its cumulative reward.
This learning has to be clearly described how it operates.
In evolutionary systems, a "fitness function" determines which agents are more likely to reproduce. This is analogous to natural selection, where organisms evolve to maximize their chances of survival and reproduction.
Some AI systems, such as nearest-neighbor, reason by analogy rather than being explicitly goal-driven. However, even these systems can have goals implicitly defined within their training data.
Such systems can still be benchmarked by framing the non-goal system as one whose "goal" is to accomplish its narrow classification task.[7]
Systems not traditionally considered agents, like knowledge-representation systems, are sometimes included in the paradigm by framing them as agents with a goal of, for example, answering questions accurately. Here, the concept of an "action" is extended to encompass the "act" of providing an answer. As a further extension, mimicry-driven systems can be framed as agents optimizing a "goal function" based on how closely the IA mimics the desired behavior.[2] In generative adversarial networks (GANs) of the 2010s, an "encoder"/"generator" component attempts to mimic and improvise human text composition. The generator tries to maximize a function representing how well it can fool an antagonistic "predictor"/"discriminator" component.[8]
While symbolic AI systems often use an explicit goal function, the paradigm also applies to neural networks and evolutionary computing. Reinforcement learning can generate intelligent agents that appear to act in ways intended to maximize a "reward function".[9]
I expect to reach that goal you must be an expert. I also expect if you want to learn you need continous feedback if you are actual learning or not. This raises the question who is actual deciding on that.
Sometimes, instead of setting the reward function directly equal to the desired benchmark evaluation function, machine learning programmers use reward shaping to initially give the machine rewards for incremental progress.[10] Yann LeCun stated in 2018, "Most of the learning algorithms that people have come up with essentially consist of minimizing some objective function."[11] AlphaZero chess had a simple objective function: +1 point for each win, and -1 point for each loss. A self-driving car's objective function would be more complex.[12] Evolutionary computing can evolve intelligent agents that appear to act in ways intended to maximize a "fitness function" influencing how many descendants each agent is allowed to leave.[4]
The mathematical formalism of AIXI was proposed as a maximally intelligent agent in this paradigm.[13] However, AIXI is uncomputable. In the real world, an IA is constrained by finite time and hardware resources, and scientists compete to produce algorithms that achieve progressively higher scores on benchmark tests with existing hardware.[14]

3. Agent function

An intelligent agent's behavior can be described mathematically by an agent function. This function determines what the agent does based on what it has seen. A percept refers to the agent's sensory inputs at a single point in time. For example, a self-driving car's percepts might include camera images, lidar data, GPS coordinates, and speed readings at a specific instant. The agent uses these percepts, and potentially its history of percepts, to decide on its next action (e.g., accelerate, brake, turn). The agent function, often denoted as f, maps the agent's entire history of percepts to an action.[15] Mathematically, this can be represented as: f:P∗→A {\displaystyle f:P^{*}\rightarrow A} Where: P\* represents the set of all possible percept sequences (the agent's entire perceptual history). The asterisk (*) indicates a sequence of zero or more percepts. A represents the set of all possible actions the agent can take.
f is the agent function that maps a percept sequence to an action.
It's crucial to distinguish between the agent function (an abstract mathematical concept) and the agent program (the concrete implementation of that function).
The agent function is a theoretical description.
The agent program is the actual code that runs on the agent. The agent program takes the current percept as input and produces an action as output. The agent function can incorporate a wide range of decision-making approaches, including:[16]
Calculating the utility (desirability) of different actions.
Using logical rules and deduction.
Employing fuzzy logic.
Other methods.
Reading the above gives me the impression that the development of any agent is a huge exercise, requiring a detailed description of all what is involved.
I would not call that a theoretical description

4. Classes of intelligent agents

4.1 Russell and Norvig's classification

4.1.1 Simple reflex agents

4.1.2 Model-based reflex agents

4.1.3 Goal-based agents

4.1.4 Utility-based agents

4.1.5 Learning agents

4.2 Weiss's classification

4.3 Other

5. Hierarchies of agents

6. Alternative definitions and uses

"Intelligent agent" is also often used as a vague term, sometimes synonymous with "virtual personal assistant".[25] Some 20th-century definitions characterize an agent as a program that aids a user or that acts on behalf of a user.[26] These examples are known as software agents, and sometimes an "intelligent software agent" (that is, a software agent with intelligence) is referred to as an "intelligent agent".
According to Nikola Kasabov in 1998, IA systems should exhibit the following characteristics:[27]
Accommodate new problem solving rules incrementally.
Adapt online and in real time.
Are able to analyze themselves in terms of behavior, error and success.
Learn and improve through interaction with the environment (embodiment).
Learn quickly from large amounts of data.
Have memory-based exemplar storage and retrieval capacities.
Have parameters to represent short- and long-term memory, age, forgetting, etc.

6.1 Agentic AI

In the context of generative artificial intelligence, AI agents (also referred to as compound AI systems) are a class of intelligent agents distinguished by their ability to operate autonomously in complex environments. Agentic AI tools prioritize decision-making over content creation and do not require human prompts or continuous oversight.[28]
They possess several key attributes, including complex goal structures, natural language interfaces, the capacity to act independently of user supervision, and the integration of software tools or planning systems. Their control flow is frequently driven by large language models (LLMs).[29]
The document "AI Agents that Matter" is not simple.
Researchers and commentators have noted that AI agents do not have a standard definition.[29][30][31][32]
Three questions arise: What is an AI agent? What is the structure? What is the programming language used.
A common application of AI agents is the automation of tasks—for example, booking travel plans based on a user's prompted request.[33][34] Prominent examples include Devin AI, AutoGPT, and SIMA.[35] Further examples of agents released since 2025 include OpenAI Operator,[36] ChatGPT Deep Research,[37] Manus,[38] Quark (based on Qwen),[39] AutoGLM Rumination,[39] and Coze (by ByteDance).[39] Frameworks for building AI agents include LangChain,[40] as well as tools such as CAMEL,[41][42] Microsoft AutoGen,[43] and OpenAI Swarm.[44]
Companies such as Google, Microsoft and Amazon Web Services have offered platforms for deploying pre-built AI agents.[45]
Proposed protocols for standardizing inter-agent communication include the Agent Protocol (by LangChain), the Model Context Protocol (by Anthropic), AGNTCY,[46] Gibberlink,[47] the Internet of Agents,[48] Agent2Agent (by Google),[49] and the Agent Network Protocol.[50] Software frameworks for addressing agent reliability include AgentSpec, ToolEmu, GuardAgent, Agentic Evaluations, and predictive models from H2O.ai.[51]
In February 2025, Hugging Face released Open Deep Research, an open source version of OpenAI Deep Research.[52] Hugging Face also released a free web browser agent, similar to OpenAI Operator.[53] Galileo AI published on Hugging Face a leadership board for agents, which ranks their performance based on their underlying LLMs.[54]

6.2 Autonomous capabilities

The Financial Times compared the autonomy of AI agents to the SAE classification of self-driving cars, comparing most applications to level 2 or level 3, with some achieving level 4 in highly specialized circumstances, and level 5 being theoretical.[55]

6.3 Multimodal AI agents

In addition to large language models (LLMs), vision language models (VLMs) and multimodal foundation models can be used as the basis for agents.
In September 2024, Allen Institute for AI released an open source vision language model, which Wired noted could give AI agents the ability to perform complex computer tasks, including the possibility of automated computer hacking.
Nvidia released a framework for developers to use VLMs, LLMs and retrieval-augmented generation for building AI agents that can analyze images and videos, including video search and video summarization.
Microsoft released a multimodal agent model - trained on images, video, software user interface interactions, and robotics data - that the company claimed can manipulate software and robots.[59]
All of this are computer applications and has nothing to do with AI.
It is important that the companies behind these AI Agents have full responsability and should expres that in writing.

6.4 Applications

6.5 Proposed benefits

6.6 Concerns

Concerns include potential issues of liability,[63][70] an increased risk of cybercrime,[33][63] ethical challenges,[63] as well as problems related to AI safety[63] and AI alignment.[33][65] Other issues involve data privacy,[33][71] weakened human oversight,[33][63][69] a lack of guaranteed repeatability,[70] reward hacking,[72] algorithmic bias,[71][73] compounding software errors,[33][35] lack of explainability of agents' decisions,[33][74] security vulnerabilities,[33][75] problems with underemployment,[73] job displacement,[34][73] and the potential for user manipulation,[74][76] misinformation[69] or malinformation.[69] They may also complicate legal frameworks and risk assessments, foster hallucinations, hinder countermeasures against rogue agents, and suffer from the lack of standardized evaluation methods.[65][77][78] They have also been criticized for being expensive[29][77] and having a negative impact on internet traffic,[77] and potentially on the environment due to high energy usage.[70][79][80] There is also the risk of increased concentration of power by political leaders, as AI agents may not question instructions in the same way that humans would.[72]
Journalists have described AI agents as part of a push by Big Tech companies to "automate everything".[81] Several CEOs of those companies have stated in early 2025 that they expect AI agents to eventually "join the workforce".[82][83] However, in a non-peer-reviewed study, Carnegie Mellon University researchers tested the behavior of agents in a simulated software company and found that none of the agents could complete a majority of the assigned tasks.[82][84] Other researchers had similar findings with Devin AI.[85]
Yoshua Bengio warned at the 2025 World Economic Forum that "all of the catastrophic scenarios with AGI or superintelligence happen if we have agents".[86]
It is important that all autonomous AI systems have a Script explaining the details and a test script. All AI systems also should have an owner.
In March 2025, Scale AI signed a contract with the United States Department of Defense to work with them, in collaboration with Anduril Industries and Microsoft, to develop and deploy AI agents for the purpose of assisting the military with "operational decision-making".[87] Researchers have expressed concerns that agents and the large language models they are based on could be biased towards aggressive foreign policy decisions.[88][89]
It is very dangerous to use LLM as part of AI systems. Part of the problem is that LLM's are not guaranteed to be stable.
Research-focused agents have the risk of consensus bias and coverage bias due to collecting information available on the public Internet.[90] NY Mag unfavorably compared the user workflow of agent-based web browsers to Amazon Alexa, which was "software talking to software, not humans talking to software pretending to be humans to use software."[91]
Agents have been linked to the dead Internet theory due to their ability to both publish and engage with online content.[92]
Agents may get stuck in infinite loops.[36][93]
Since many inter-agent protocols are being developed by large technology companies, there are concerns that those companies could use these protocols for self-benefit.
The general impression is that there are many problems related with AI Agents.

6.6.1 Possible mitigation

7. Applications

The concept of agent-based modeling for self-driving cars was discussed as early as 2003.[96]
Hallerbach et al. explored the use of agent-based approaches for developing and validating automated driving systems. Their method involved a digital twin of the vehicle under test and microscopic traffic simulations using independent agents.[97]
Waymo developed a multi-agent simulation environment called Carcraft, to test algorithms for self-driving cars.[98][99] This system simulates interactions between human drivers, pedestrians, and automated vehicles. Artificial agents replicate human behavior using real-world data.
Salesforce's Agentforce is an agentic AI platform that allows for the building of autonomous agents to perform tasks.[100][101]
The Transport Security Administration is integrating agentic AI into new technologies, including machines to authenticate passenger identities using biometrics and photos, and also for incident response.[102]

8. See also

Following is a list with "Comments in Wikipedia" about related subjects

Reflection 1 - AI agent versus intelligent agent

The general impression of this document is the overestimation of AI agents in science.
It should be understood that there is a difference between Intelligence and Artificial Intelligence.
Intelligence is a human capability. A human performs a intelligent action if he or she is capable to solve a new problem, only by discussions with other humans or by reading books written by humans. That means always more humans are involved, that means non of the humans separately, know the answer. The same with reading books or a combination. That means an intelligent human finds the solution by using and understanding different resources intelligently.
Artificial Intelligence is functional the same Intelligence, except that the human aspect is replaced by a computer program. That means Artificial Intelligence is a computer program which finds the solution by using and understanding different resources intelligently. The resources in this case are discussions with humans and reading books.
The emphasis is a computer program. The chalenge is that it is very difficult to write such a program, which is capable to understand text and to raise questions.

An Intelligent agent is a robot (automobile) and a computer equipped with sensors and activators. An Intelligent agent performs a fixed number of task and requires updates in order to implement new functionality.

Reflection 2

Reflection 3

Feedback

If you want to give a comment you can use the following form Comment form
Created: 3 September 2024

Go Back to Wikipedia Comments in Wikipedia documents
Back to my home page Index