Practical Realities: How AI Empowers Quantum Technologies to Solve Today’s Biggest Challenges

By Fernando Dominguez Pinuaga, Andrea Bortolato, Alfonso Mu?oz Mu?oz and Stefan Bogdanovic

At its core, AI is about solving complex problems efficiently. This efficiency is derived from its ability to analyze vast amounts of data, recognize patterns, and increase productivity. The rise of AI today has three fundamental roots:

1. The quality and size of the models developed, federated AI, AutoML, reinforcement learning, large language models (LLMs) and others?
2. The vast amount of data we generate that allows us to train those models?
3. The increase in hardware capabilities including GPUs that enable us to run those models at scale.

Advances in all these three areas are driving applications such as image recognition, content generation, self-driving cars, or summarizing thousands of customers’ reviews.? What is less well known is that AI is playing a pivotal role in advancing quantum technologies.?

Quantum physics, with its roots going back a century, introduces a complete shift in our understanding of the physical world. We learn of a wondrous hidden complexity underlying all phenomena, and within that complexity is the potential for a number of breakthrough applications. However, harnessing the complexity is easier said than done. Even though many modern technologies only exist because of quantum physics, the truth is that we have only scratched the surface of what is possible.? You can find great videos on quantum tech in our YouTube channel.?

You have used quantum tech without even knowing it. Examples of quantum-driven technologies that are all around us include: lasers (including the laser pointer!), the MRI scanner and the transistor. Now a new generation of quantum tech is emerging and it is combining with AI to drive even faster to market.????

?Let’s look at some examples:

Simulation

The optimal combination of molecular simulation and AI helps accelerate R&D by honing in on positive potential outcomes much faster and more efficiently than exclusively using experimental data, delivering faster time to insights and more efficient, commercially-relevant outcomes. By simulating the quantum interactions between molecules – such as a drug compound and target receptors in the human body – we generate new data that AI models can use to predict the success of said molecules.

For instance, in biochemistry, simulation accelerates drug discovery by greatly reducing the time, cost and risk of testing of new drug compounds. With simulation, researchers can run billions of simulations in silico, simultaneously testing for efficacy and toxicity and adjusting their molecular structure much faster than they could ever do via lab experiments. AI is already having a tremendous impact on the healthcare industry with its ability to analyze vast stores of data.??

In the case of challenging medical conditions such as Alzheimer’s or cancer, where available data are not enough for impactful AI solutions, quantum mechanics starts where AI ends. Researchers are using ab initio data generated by quantum simulation to power AI models, leading to breakthroughs in treatments that have stymied the industry for years and shaving several years and millions of dollars from their R&D efforts.

Similar advances are being made in materials science, which leverages simulation to create new chemical compounds, materials or products that are better, stronger, safer, more durable, cost effective, sustainable and eco-friendly. Simulation will have a significant impact across numerous industries, such as aerospace, agriculture, chemicals, construction, defense, energy, manufacturing, and more.?

For example, the increased demand for electric vehicles (EVs), renewable energy and consumer electronics has increased global demand for lithium-ion batteries. Traditionally, it takes years to physically test the performance and life span of each new battery design in the lab. Simulation can dramatically reduce that time to weeks or months, allowing battery manufacturers to focus on only the most promising formulas and designs, which reduces R&D time, risk and costs, while accelerating the introduction of new products to the market.

On the sustainability front, researchers can efficiently explore a vast array of material compositions and properties without the need for extensive physical tests,? accelerating the discovery of new materials that might be more effective, easier to procure, and less harmful to the environment. Additionally they can use it to reverse some of the damage we’ve already caused, such as developing water soluble plastics or breaking the molecular bonds of so-called "forever chemicals.”

Recently, companies such as NVIDIA have taken an interest in advancing simulation for materials and life science applications, as it requires the computing power of GPUs. Working with partners like SandboxAQ, the company is actively improving its hardware and software capabilities to usher in a new era in computing centered around simulation and AI.

Sensing

Quantum sensors leverage the inherent fragility of quantum states to detect minute changes in the physical world, such as gravity, magnetism, acceleration. When AI is applied to the plethora of data generated by these highly sensitive sensors, researchers can isolate and extract the signals required to run various applications.? Below are three ways AI unlocks the power of quantum sensing

• Noise Mitigation: Due to their unparalleled sensitivity, quantum sensors often face strong interference from stray environmental fields (e.g., electromagnetism, gravity, vibration, etc.), which limits their effectiveness. Traditional solutions, for example, an MRI machine,? prevent interference by building complex shielding structures that are large, costly and impractical. AI, however, excels in distinguishing useful signals from noise, employing algorithms to efficiently filter and clean up the desired signals, thereby improving fidelity and mitigating interference. It also enables us to build solutions that are smaller, less complex and require fewer sensors to achieve the same signal-to-noise performance as larger, more complex systems that don’t use AI.

• Deriving Actionable Insights: The complexity and volume of data generated by the novel quantum sensors can overwhelm traditional analysis methods. AI algorithms are instrumental in parsing extensive data streams, detecting patterns in complex data and classifying quantum sensor readings. At SandboxAQ, we use these algorithms to transform intricate magnetic field maps of human heart activity into actionable insights for cardiovascular diagnosis. As a second application, we utilize quantum sensors to read Earth's magnetic field and employ AI to precisely match an aircraft’s location to known coordinates of the magnetic map (and remove the interference generated by the aircraft itself) as a robust supplement to GPS navigation.

• Sensor Development: Designing and building high performing sensors and multi-sensor systems requires exploring enormous and complex parameter space, which is intractable by humans. AI enables designers to efficiently traverse this multitude of design possibilities, incorporate real-world constraints, rapidly iterate prototypes and deliver an optimized sensor system design with superior capabilities.

There are many other use-cases for quantum sensing across industries. Biophysical sensors could be used in wearable medical devices for continuous monitoring, or improved human/machine interfaces. Geophysical sensors could be used for underwater or underground exploration and early earthquake detection. In the tech realm, they could be used for hard drive diagnostics, semiconductor imaging or chip-failure analysis, lithium-ion battery fault detection and more.?

Communications and Cryptography

AI contributes significantly to enhancing cryptography and privacy and enabling quantum communications.?

AI can process large amounts of data, analyze network traffic, recognize patterns and trends, and monitor for abnormalities much faster than humans, enabling real-time threat detection, analysis and remediation. It can also flag future vulnerabilities and guide the development of more secure cryptographic methods. AI brings real and actionable value to an organization’s cybersecurity, mitigating organizational cybersecurity risks using frameworks and guidelines such as NIST’s cybersecurity Framework, MITRE ATT&CK or zero-trust methodologies.

In addition to the above, it is becoming increasingly important that security teams be able to analyze and work with encrypted data. In this scenario, AI is vital to develop privacy-enhancing technologies that minimize the use of personal data and maximize data security. For example, developing AI models that can detect misconfigurations of encrypted data captured from a network or an organization while preserving the privacy of other user information.

Secure network communications is another scenario where AI plays a major role, helping us harness advanced quantum physics principles like entanglement and superposition to create quantum communication networks. These networks create quantum states that collapse if a third-party tries to observe them or hack into the data being transmitted. The communications are carried out by various physical means (fiber-optic cable, air, terrestrial or satellite) and usually require specialized hardware and software on top of communication protocols of varying complexity. AI is useful in process optimization and fault elimination.

Finally, in quantum computing, new AI techniques allow us to correct the way qubits interact with each other or help design chips in a more efficient manner. Both of these use-cases help accelerate the development of fault tolerant quantum computers by allowing engineers to discover new methodologies faster.?

Thanks to the advances AI technologies have made since the turn of the century, harnessing quantum physics has evolved from a science problem to an engineering challenge. However, we never expected AI’s incredible progress would make engineering so much more efficient. The synergy of AI and quantum physics represents a new frontier of innovation with far-reaching implications across a broad range of industries and business use-cases. As we tap into the power of AQ, we not only unravel the mysteries of quantum mechanics but also pave the way for unprecedented technological transformation and groundbreaking applications in simulation, sensing, and secure communication.?