Quantum Computing
The Limits of Classical Computing Power Broken
Although current advances in the field of AI are impressive, we are slowly running up against the physical limits of traditional computers. Moore’s Law (which predicts that the computing power of microchips doubles every two years) is reaching the point where transistors are becoming so small that quantum effects disrupt their operation. To make the next giant leap in machine learning and data science, the IT industry is looking to a fundamentally new architecture: Quantum Computing. The combination of Quantum Computing and AI — also known as Quantum Machine Learning (QML) — promises to solve complex problems that would take our current supercomputers billions of years to solve.
Although we are still in the pioneering phase (the ‘NISQ era’), tech giants like IBM, Google, Microsoft, and specialized startups are investing billions in building stable quantum computers. The impact on sectors such as logistics, materials science, and cryptography will be revolutionary.
Qubits: Superposition and Entanglement
To understand the potential of Quantum AI, we must look at the basic principles. Classical computers work with bits that have the value 0 or 1. Everything you currently see on your screen is the result of billions of ones and zeros calculated sequentially. Quantum computers use ‘qubits’. Thanks to the quantum mechanical phenomenon called ‘superposition’, a qubit can assume 0, 1, or any value in between *simultaneously*.
Additionally, qubits can become ‘entangled’ with one another. This means that the state of one qubit is directly connected to the state of another, regardless of the distance between them. These properties enable a quantum computer to perform not one calculation at a time, but to calculate all possible outcomes of a complex problem in parallel. For AI models that need to optimize enormous multi-dimensional datasets, this is the holy grail.
Quantum Machine Learning: Unrivaled Optimization
Machine learning essentially revolves around optimization: finding the lowest error rate in a vast landscape of mathematical possibilities (gradient descent). For traditional computers, training a model like GPT-4 is a process that takes months on tens of thousands of GPUs, at the cost of enormous amounts of energy. Quantum algorithms have the theoretical potential to shorten these training processes from months to hours or even minutes.
Moreover, quantum computers are superior at recognizing patterns in extremely noisy, unstructured data. This makes them ideal for the pharmaceutical industry. Instead of testing millions of chemical combinations in laboratories for a new drug, a quantum AI model can simulate the behavior of complex protein structures and molecules at the subatomic level, which will drastically accelerate the discovery of life-saving medicines.
The Black Swan for Cybersecurity
The rise of quantum computers also poses an enormous threat, specifically in the field of cybersecurity. The encryption currently securing the entire internet (such as RSA and ECC, used for HTTPS, banking transactions, and VPNs) relies on mathematical problems — such as factoring enormous prime numbers — that are virtually impossible for classical computers to solve. However, with the so-called ‘Shor algorithm’, a powerful quantum computer can crack this encryption within a few hours.
This has led to an arms race around Post-Quantum Cryptography (PQC). Governments and IT leaders are urging organizations to start inventorying and migrating their encryption standards now. Malicious actors are already engaged in the mass interception and storage of encrypted data traffic (‘harvest now, decrypt later’), with the aim of decrypting this data as soon as quantum computers become commercially available.
Preparing for the Quantum Age
For IT professionals and business leaders, quantum computing might still sound like science fiction, but the adoption curve will be steeper than we expect. Soon, companies will not need to build a physical quantum computer in their basement; these systems will be accessible via cloud services (Quantum-as-a-Service), such as Amazon Braket or Microsoft Azure Quantum.
The task for innovative organizations is to now experiment with quantum simulators and hybrid architectures, where classical cloud environments collaborate with quantum processors for specific, heavy optimization tasks. Those who are the first to harness the power of Quantum Machine Learning will gain an insurmountable competitive advantage in their markets.
Delve into the impact of this fundamental technological shift on your business strategy through this informative resources about Quantum IT on AG Connect.
AI Governance and the AI Act: Navigating Ethics and Legislation
