
Quantum AI
What it is and why it matters
Quantum AI combines the power of quantum computing with artificial intelligence. This integration plays on the unique strengths of quantum and AI, using quantum bits, known as qubits, to perform advanced computations that classical computers can’t handle.
History of quantum AI
The concept of quantum computing emerged in the early 1980s when physicist Richard Feynman proposed the idea of using quantum mechanics to simulate physical systems, which classical computers could not do. This idea laid the foundation for quantum computing, which uses the principles of quantum mechanics, including superposition and entanglement, to perform complex computations.
In the 1990s, the development of quantum algorithms, such as Shor's algorithm for factoring large numbers, demonstrated the potential of quantum computing to solve problems faster than classical computers. These advancements spurred interest in exploring the intersection of quantum computing and AI.
The early 2000s saw the establishment of the Quantum Artificial Intelligence Lab by NASA, Google and the Universities Space Research Association. This initiative aimed to pioneer research on how quantum computing could enhance machine learning and other complex computational tasks.
Around the same time, researchers began developing quantum machine learning algorithms, which leverage quantum computing to improve the speed and accuracy of AI models.
In recent years, the focus has shifted toward practical applications of quantum AI.
Companies at the forefront of this research are exploring hybrid architectures that combine quantum and classical computing. For instance, current research investigates the use of quantum annealing for optimization problems and the gate model for more universal applications like machine learning, quantum chemistry and simulation.
Quantum AI in today’s world
Quantum AI explained
Learn about quantum computing and how it works from Amy Stout, Head of Quantum AI Product Strategy at SAS. What is a qubit? And how does quantum computing differ from classical computing? Stout provides clear answers and explains where we're headed with this emerging technology.
Who's using quantum AI
Quantum AI has the potential to revolutionize industries by providing unprecedented computational power and efficiency.
Consider these industries where quantum computing could make a significant impact:
How quantum AI works
Quantum computers are different from any existing classical computer, including smartphones and even the most powerful supercomputers. They take advantage of the unique properties of quantum mechanics, such as superposition and entanglement, to help solve certain classes of complex problems that are too challenging for classical computers to solve alone. In some cases, they can solve the problem significantly faster, and in other cases, they can represent the problem in ways that conventional computers cannot.
For now, quantum computers will not replace conventional computers but work alongside them as another tool. Under this paradigm, CPUs, GPUs and QPUs will work together to address the pieces of the problem for which they are best suited.
Classical computers use bits to represent data as either 0 or 1. However, quantum computers use quantum bits, or qubits, which can exist in multiple states simultaneously. The principle of superposition describes the existence of multiple states and can be illustrated with the following analogy:
Consider a coin. There are two clear states the coin can exist in, heads or tails, which can be thought of as the zero and one state of a classical bit. Now imagine the coin is spinning in the air. In this case, the heads and tails states exist together with an equal probability of measuring either state once the coin stops. Quantum computing can use this simultaneous nature by performing calculations on both heads (0) and tails (1) at the same time, as long as the coin remains spinning (in a state of superposition).
This state of superposition allows quantum computers to process double the amount of information in a single qubit vs. a single bit. As you increase the number of qubits, the amount of information that can be processed increases exponentially as 2number of qubits, significantly speeding up computations. For example, 10 qubits can perform the calculations equivalent to 1,024 classical bits, and this grows exponentially.
Next, let's learn about entanglement and quantum algorithms:
Entanglement
Another equally important quantum physical property used in quantum computing is entanglement. You can simply think of entanglement as quantum particles being correlated. When two qubits are entangled, if you know the state of one, then you automatically know the state of the other. Entanglement, when combined with superposition, can further increase computational power.
Quantum algorithms
Quantum AI also uses quantum algorithms to improve machine learning models. Quantum machine learning algorithms, such as quantum support vector machines and quantum neural networks, use quantum circuits to perform computations.
These quantum circuits represent a universal method of performing quantum computations.
For instance, in a common implementation of a quantum neural network, classical data is encoded into quantum states. The quantum circuit uses parameterized rotations, entanglements and measurements to examine complex relationships simultaneously. The output is classically optimized and fed back into the circuit as new parameterized rotations, repeating until an optimal configuration is derived. This is similar to optimizing node weights in a classical neural network.
Quantum AI – a hybrid approach
Since quantum computing technology is still maturing, quantum AI is a hybrid process involving a combination of quantum and classical computing approaches. In some cases, quantum processing happens first, in other cases it happens last, and sometimes there is a cyclical approach between quantum and classical computing. This hybrid nature uses the strengths of both quantum and classical computing to achieve better performance and accuracy.
As quantum computers evolve, we will continue to see hybrid approaches that use the increasing reliability and scalability of quantum computers to enhance AI-assisted decisioning.
We are at the dawn of the integration between quantum computing and AI. This integration will become tighter as quantum computing matures. Currently, quantum computing manufacturers are experimenting with co-locating QPUs (quantum processing units) and specialized AIUs (AI units).
As this type of research evolves over the next 5 to 10 years, expect to see tremendous technological quantum-AI gains. These advances will change our existing methodologies and open doors to help solve complex problems in new and unique ways.
Recommended reading
-
Article AI transforms insurance: See 8 examples of how it worksUsing AI in insurance can boost value for customers, insurance companies and stakeholders alike. Examples range from stronger fraud detection and improved customer service to optimized underwriting processes and competitive advantage. Learn more about how AI is transforming the industry.
-
Article What is AI modeling?AI modeling involves creating programs that use one or a combination of algorithms to allow computers to think, learn and predict outcomes. Much like a human brain, AI models absorb input data – numbers, texts, images, video, sound – to learn how to predict outcomes or solve specific tasks without explicit instructions at every step.
-
Insights Page Understanding digital twin technologyLearn how digital twin technology can change industries from health care to manufacturing, and shape the future of data modeling and process improvements.