DEV Community

24P-0505
24P-0505

Posted on

AI

Artificial Intelligence research continues to evolve rapidly, but many modern systems still rely on fundamental ideas taught in AI courses. Two recent research papers — The Rise of Agentic AI: A Review of Definitions, Frameworks, and Challenges (2025) and Research on the A* Algorithm Based on Adaptive Weights (2025) — demonstrate how classical AI concepts such as intelligent agents and search algorithms remain central to modern developments.

This blog discusses the goals of these papers and connects their ideas to topics studied in Artificial Intelligence, particularly agent architectures and heuristic search methods.

Paper 1: The Rise of Agentic AI

Goal of the Paper

The first paper explains the concept of Agentic AI, which refers to AI systems that can autonomously perform tasks to achieve specific goals. Traditional AI models mainly generate responses based on prompts, such as answering questions or producing text. Agentic AI systems go further by planning tasks, making decisions, and interacting with tools or environments.

For example, an AI research assistant might search for information, analyze the results, generate new queries, and combine the information into a final response. This ability to perform multi-step reasoning and actions makes agentic AI more capable than traditional AI systems.

Connection to AI Course Concepts

The idea of agentic AI relates directly to the Intelligent Agents topic studied in AI courses. In AI theory, an agent perceives its environment, processes information, and takes actions to achieve a goal.

Agentic AI follows this model but integrates technologies such as large language models, APIs, and external tools. These systems behave similarly to goal-based agents, where the agent evaluates possible actions and selects the best option to reach the desired outcome.

This research shows that agent architectures discussed in AI textbooks are now applied in real AI systems, including AI assistants, research tools, and automated task planners.

Paper 2: Research on the A* Algorithm Based on Adaptive Weights

Goal of the Paper

The second research paper focuses on improving the A* search algorithm, which is widely used for pathfinding and optimization problems.

The traditional A* algorithm uses the formula:

f(n) = g(n) + h(n)

Here, g(n) represents the cost from the start node to the current node, while h(n) represents the heuristic estimate from the current node to the goal. The algorithm expands the node with the lowest cost until the goal is reached.

The research introduces adaptive weighting, where the influence of the heuristic changes during the search. This allows the algorithm to balance search speed and accuracy more effectively.

Connection to AI Course Concepts

This research relates directly to heuristic search methods studied in AI courses. Common algorithms include Breadth First Search (BFS), Depth First Search (DFS), Best First Search, and A* Search.

Adaptive weighting improves A* by guiding the search more efficiently in large problem spaces. This approach is useful in applications such as robot navigation, game AI pathfinding, GPS routing, and autonomous vehicles.

Personal Insights from Reading the Papers

Reading the papers manually and exploring them with NotebookLM provided different perspectives. Manual reading helped understand the technical aspects, such as how adaptive weights influence the heuristic function and how agent systems structure decision making.

NotebookLM helped summarize complex sections and highlight key ideas more quickly. It made it easier to identify the main contributions of the papers and connect them with AI concepts such as agents, heuristics, and search strategies.

These papers show that even with rapid technological advances, modern AI systems still rely on classical AI foundations like search algorithms and agent-based models.

Top comments (0)