Recently, it seems that it has become fashionable to talk about artificial intelligence (AI), and this expression is now used wherever possible. But, despite this, the technology itself can be useful in several areas at once. Similarly, quantum computing has received renewed attention as a purported revolutionary tool that could, among other things, enhance cyber defenses and even create a new Internet. And although in recent years both technologies have seriously advanced, they are still far from perfect, no matter how anyone would like it to be otherwise.
This is especially true for AI, which in its current form is mainly specialized machine learning algorithms that can automatically perform individual tasks. According to researchers at the Center for Quantum Technology at the National University of Singapore (NUS), AI can be significantly improved through quantum computing.
In a new study published in the journal Physical Review Letters, scientists at NUS have proposed a quantum algorithm for linear systems of equations that will enable much more efficient quantum computer analysis of large datasets.
“Previously similar quantum algorithms were used for a narrow range of problems. We have to improve them if we want to achieve quantum acceleration for other data,”study author Zhao Zhikuan said in a press release.
A quantum algorithm, in simple terms, is an algorithm designed to run within realistic quantum computational models. Like traditional algorithms, quantum is a step-by-step procedure, but it uses phenomena specific to quantum computing, such as quantum entanglement and superposition.
In this case, the algorithm for solving linear systems performs calculations using large data matrices. Such large-scale tasks are more suitable for quantum computers.
Better, faster, stronger
Promotional video:
In other words, the algorithm for solving linear systems offers a faster and more powerful way of computing compared to classical computers. The first version of such a quantum algorithm, developed in 2009, laid the foundation for research into quantum forms of AI and machine learning.
"Quantum machine learning is an evolving field where researchers are trying to harness the power of quantum information processing to accelerate the execution of classic machine learning tasks," the research paper says. Whether this will make AI smarter is another question.
Today's AI systems and their machine learning algorithms are already capable of performing enormous amounts of computation. The process of processing datasets (and this is usually tons of information through which the AI makes its way) will definitely be accelerated by quantum computing.
Of course, before the algorithm developed by Zhao and his colleagues can be useful, we need to create more suitable quantum computers. Given the amount of work done on this front, it can be assumed that it won't be long before the concept becomes a reality.
“We anticipate that it will take three to four years for current hardware experiments to become real-world applications for quantum computing in artificial intelligence,” Zhao said in a press release. In the meantime, his team plans to conduct a demonstration of how their algorithm works soon.
Dmitry Volkov