Digital News

　　What new landscape might emerge from the integration of artificial intelligence (AI) and quantum computing, two of the most high-profile frontiers of future technology? Days ago, at a recent special session of the Global Artificial Intelligence Technology Conference 2025 (GAITC 2025) held in Hangzhou, industry experts noted that “Quantum Computing and AI Integration” is poised to drive a new wave of technological innovation.

　　“The core of Quantum Computing and AI Integration is not simply the superposition of technologies, but a horizontal convergence across multiple domains and disciplines, achieving nonlinear growth through collaborative innovation”, said Academician Dai Qionghai of the Chinese Academy of Engineering. On the one hand, quantum computing may break through the computational bottlenecks currently hindering AI model training and improve algorithmic efficiency. On the other hand, AI can empower quantum technologies in reverse, enhancing quantum control, error correction, and algorithm design, thereby providing new paths for the stability and scalability of quantum systems.

　　“In the past five years, breakthroughs in AI, especially generative AI, have upended many computing paradigms. In the next five years, quantum computing is likely to move from the lab to real-world applications. So, the convergence of AI and quantum computing is likely to become an inevitable trend”, said Sun Xiaoming, a researcher at the Institute of Computing Technology, Chinese Academy of Sciences.

　　Professor Long Guilu of Tsinghua University and Vice President of the Beijing Academy of Quantum Information Sciences pointed out that there are currently two major directions in the integration of quantum information and AI: first, AI for Science, such as Google’s use of quantum AI to optimize error correction codes last year; and second, quantum computing empowering machine learning, where mature quantum computers could provide computation power to support AI in the future.

　　Industry, academia, and research institutions are actively exploring innovations in Quantum Computing and AI Integration and have already achieved notable progress. For instance, the team of Lu Liqiang, a Hundred-Talent Program Researcher at the College of Computer Science and Technology, Zhejiang University, used a Mixture of Experts (MoE) model to improve calibration quality, boosting quantum state distinguishability by 25.5%. Meanwhile, their waveform optimization technology based on convolutional matching increased quantum circuit compilation speed by 158 times. At Shanghai Jiao Tong University, Xiao Tailong’s team was the first to apply quantum machine learning to single-pixel imaging systems, overcoming the traditional reliance on large labeled datasets, and experimentally demonstrating the advantage of quantum feature space in extracting information under low sampling rates. The team at the Beijing Academy of Quantum Information Sciences has also achieved recent breakthroughs in quantum node embedding algorithms, quantum convolutional neural networks, and quantum resonance-based dimensionality reduction algorithms. Beijing Boson Quantum Technology Co., Ltd. (hereinafter referred to as “Boson Quantum”), a company based in Beijing, proposed a quantum training method based on coherent light quantum computers, replacing traditional Gibbs sampling with quantum sampling, significantly improving the training efficiency of Boltzmann machines. The company also collaborated with the Guangzhou National Laboratory to develop a quantum algorithm for protein structure prediction, overcoming challenges that traditional algorithms struggled to address in complex scenarios.

　　“Quantum computing has vast application potential in pharmaceuticals, finance, and AI-driven manufacturing. In drug discovery, in particular, a hybrid ‘Quantum Computing + AI’ approach can efficiently screen target-specific molecules from vast compound spaces, significantly reducing R&D costs and timelines”, said Gao Qi, R&D Director at Boson Quantum.

　　More and more cities are joining the race to explore Quantum Computing and AI Integration as a new track for future industries. Currently, cities like Hangzhou and Hefei are promoting the integration of quantum computing into the AI ecosystem and accelerating the implementation of “Quantum Computing + AI” technologies.

　　“We deeply recognize that AI is the foundational layer, quantum technology is the driving force for breakthroughs, and their integration is the key path to seize the lead in future industries and global discourse power,” said Li Bo, a member of the Party Working Committee of the Hangzhou Future Sci-Tech City Management Committee. He added that Hangzhou has established a “1+3+X” future industry system, with AI as the base and a focus on three emerging fields: low-altitude economy, humanoid robotics, and brain-inspired intelligence, while proactively deploying cutting-edge areas like quantum information as the “X” component. “In the future, we will prioritize quantum algorithms to accelerate AI training, neuromorphic computing, and other integrated tracks, aiming to bring more breakthroughs from zero to one.”

　　However, many experts believe there are still significant challenges in the field of Quantum Computing and AI Integration that must be addressed. Xiong Hongkai, a Specially-appointed Professor at Shanghai Jiao Tong University, pointed out that if AI models are directly optimized using quantum computing, the powerful computational capability of quantum computers can help realize fully quantized models. Current efforts in optical and quantum computing are exploring this direction, but both face their constraints and limitations.

　　Lu Liqiang also noted that despite challenges such as limited qubit counts and the absence of mature theoretical paradigms, the mutual empowerment between AI and quantum has already opened up a new landscape for Quantum Computing and AI Integration. Full-stack research in areas such as chip architecture and compilation optimization is now propelling this transformation from lab experiments to real-world applications.