From Chat to Act: Quantum Computing Power Drives the Rise of AI Agents

At the 2026 Consumer Electronics Show (CES), a brand - new "CES Foundry" zone made its debut. Here, the integration of artificial intelligence and quantum computing began to showcase its potential computing power to reshape the world.

As the competition in large language models (LLMs) heats up, the industry has started to realize that the simple stacking of parameters has reached a bottleneck. The real next - generation AI consists of AI agents capable of autonomous decision - making and solving extremely complex problems. And this requires a more powerful engine.

The new book The Rise of AI Agents by CITIC Publishing Group reveals this core proposition: how quantum computing will become the "nuclear power" for AI agents, driving a crucial leap for AI from passive Chat (dialogue generation) to active Act (autonomous action).

I. The Dimensionality Reduction Strike from the Perspective of Quantum Computing

For AI agents to operate in the dynamic and complex real world, the required computational complexity far exceeds today's imagination. The key to understanding evolution lies in understanding the fundamental differences between quantum computing and the classical computing we are accustomed to.

1. Quantum Superposition

The operating unit of a classical computer is the bit, which can only be either 0 or 1. It's like walking through a maze. A classical computer has to try one path, return when it hits a dead - end, and then try another, repeating this process until it finds the exit.

The core of quantum computing is the qubit. Quantum superposition means that a quantum system can be in multiple possible states simultaneously when observed.

Based on the principle of quantum superposition, a qubit can be in the state of 0 and 1 simultaneously, or any combination of these two states. This means that a quantum computer doesn't need to process data sequentially like a classical computer.

For AI agents, this represents a qualitative change. With qubits in a superposition state, a quantum computer can explore multiple paths simultaneously, significantly reducing the time needed to find the exit.

With the support of quantum computing, in the future, when AI agents make supply - chain decisions or conduct financial risk control, they don't need to simulate countless possibilities one by one. Instead, they can use the quantum superposition state to calculate all possible scenarios simultaneously. This ability opens the door to processing massive amounts of data simultaneously, completely overturning the sequential data - processing mode of classical computing.

2. Quantum Entanglement

When two or more qubits become entangled, the state of one qubit will instantaneously affect the other, regardless of the distance between them.

This characteristic leads to quantum parallelism. If classical computing is like reading one word at a time, quantum computing is like reading several pages of a book at once. This parallel - processing ability is the foundation for future AI agents to solve extremely complex problems, such as simulating a complete ecosystem or optimizing the transportation network of a megacity in real - time.

3. Quantum Supremacy

"Quantum Supremacy" refers to the situation where a quantum computer outperforms today's most powerful classical supercomputers in processing specific tasks.

Google's Sycamore quantum processor achieved quantum supremacy in 2019. It completed a specific calculation in just 200 seconds, while the most advanced supercomputer at that time would have taken about 10,000 years to complete the same task.

For AI agents, the achievement of quantum supremacy means that they will no longer be restricted by the computing - power bottleneck and can undertake tasks that currently seem out of reach, standing at the forefront of solving cutting - edge problems in science, finance, and even broader fields.

II. The Future Defined by Quantum Algorithms

For AI agents to move from theory to practice, they must master the tools of efficient algorithms. Currently, there are two major foundational quantum algorithms. They pave the way for the rise of AI agents in terms of security and data processing respectively, while also bringing challenges.

1. Shor's Algorithm: The Sword of Damocles Hanging Over the Encryption System

In the future where AI agents are interconnected, security is the cornerstone of trust. However, the famous Shor's algorithm proves that a quantum computer can perform integer factorization exponentially faster than the best - known classical algorithm.

Current Internet security protocols (such as RSA encryption) rely on the difficulty of factoring large numbers. The emergence of Shor's algorithm means that a quantum computer capable of running this algorithm efficiently will render existing encryption systems vulnerable.

This forces AI agents to evolve - not only to utilize the power of quantum computing but also to build a post - quantum cryptography defense system.

AI agents must keep up with the times and integrate and adopt post - quantum computing encryption methods. In a future environment where quantum computing is everywhere, this is crucial for ensuring the secure operation of AI agents.

2. Grover's Algorithm: Precision "Needle - Finding" in the Ocean of Data

If Shor's algorithm is a spear, Grover's algorithm is the "eye" of AI agents.

When searching for specific information in an unstructured database (such as a vast amount of unstructured text and image data), a classical algorithm requires an average of N/2 operations, while Grover's algorithm, using quantum mechanics principles, can complete the task in approximately √10 operations.

Grover's algorithm is crucial for AI agents in the big - data era. It can not only greatly improve database retrieval efficiency but also be used to handle various pattern - recognition and optimization challenges. This means that when facing the chaotic and complex real - world data, AI agents can accurately locate key information at an astonishing speed and make wise decisions.

III. Quantum Computing Detonates AI Agents

The combination of quantum computing and artificial intelligence is not a simple addition but an exponential explosion. This integration will endow AI agents with unprecedented "action capabilities".

1. The Qualitative Change from "Big Data" to "Quantum Data"

Traditional machine learning relies on training with massive amounts of data. As the model parameters and data scale increase, classical computers face a serious computing - power bottleneck. Quantum computing, with its ability to efficiently process large - scale complex data sets, can break through this limitation. Quantum - enhanced machine - learning algorithms will be more noise - resistant, faster, and more accurate.

The ability of quantum algorithms to process massive data sets can significantly enhance the capabilities of machine - learning models. This synergy will completely change fields such as climate modeling.

2. Disruptive Applications in Vertical Fields

With the support of quantum computing, the "action" capabilities of AI agents will take root in specific industries:

Drug Discovery and Life Sciences: This is the most exciting battlefield for quantum AI. Classical computers cannot accurately simulate complex molecular structures, while quantum simulation can simulate molecules and chemical reactions with extremely high precision.

Using IBM's quantum computer to simulate the molecular structure of a small - molecule protein is a task that traditional computing methods cannot achieve. In the future, AI agents will be able to design new drugs autonomously, significantly shortening the R & D cycle and reducing costs.

Finance and Risk Management: In the financial field, AI agents can use tools such as Grover's algorithm to greatly improve the level of risk assessment and optimize complex investment portfolios in real - time.

Logistics and Supply Chain: Facing the complex global supply - chain network, quantum AI agents can solve large - scale optimization problems that classical computing power cannot handle and plan truly globally optimal paths.

IV. The Ultimate Form of AI Agents

The ultimate goal of AI development is not just text generation (Chat) but the ability to remember, perceive, reason, and act autonomously (Act) in a complex environment. Quantum computing is the key piece of the puzzle to achieve this leap.

Although the prospects of quantum computing are broad, quantum AI will not completely replace classical computing but enhance it.

The future computing architecture will be in a "hybrid mode": Classical computers will continue to be responsible for general tasks, daily data processing, and logical control. Quantum computers, like "special forces", will specialize in handling highly complex optimization tasks, molecular simulations, or cracking codes. AI agents, as "commanders", will schedule between classical and quantum resources, integrating the advantages of both to solve problems.

The integration of AI and quantum is two - way.

Not only is quantum computing accelerating AI, but AI is also helping quantum computing to be implemented. The latest research shows that artificial - intelligence technology can assist in the development and optimization of quantum algorithms. Through automated programming, it can significantly lower the development threshold of quantum computing. For example, using machine - learning models (such as diffusion models) to generate flexible and precise quantum circuits can make quantum computers more adaptable to different hardware configurations.

This cycle of "AI designing quantum, and quantum enhancing AI" may accelerate the arrival of the "technological singularity".

Original Book Title: The Rise of AI Agents: Integrating AI, Blockchain Technologies, and Quantum Computing

Note: The cover image is AI - generated

This article is from the WeChat official account "Big Data Digest". Author: [UK] Petar Radanliev. Translated by Dong Shimin and Xu Shenghui. Republished by 36Kr with permission.