Dr. Pravir Malik is the founder and chief technologist of QIQuantum and the Forbes Technology Council group leader for Quantum Computing.
getty
Substantial strides have been made in AI and quantum technology over the last five years. This article will highlight some recent developments in each area, examine the inevitable convergence already being foreshadowed by some leading players, suggest a boundary toward which the industry may be proceeding and conclude with some implications.
Recent Developments in AI
Leaps in natural language processing have driven recent developments in AI that continue to fire the imagination. These include:
• The emergence of the transformer architecture, which allows for parallel processing of sequential data and enables models to capture long-range dependencies in language efficiently.
• Scaling up the model size to 1 trillion parameters enables the processing of vastly more information to improve language understanding significantly.
• Multimodality enables AI to understand and generate content across various modalities (text, images, audio), unlocking richer and more human-like interactions.
• Unsupervised pre-training on massive text corpora, allowing LLMs to learn general language patterns and world knowledge from unlabeled data.
Recent Developments In Quantum Technology
In quantum technology, there continue to be rapid developments in a range of sensing capabilities that allow us to dive deeper into understanding nature and continue to develop a stream of powerful technological applications. These include developments in:
• Quantum magnetometry exploits defects in diamond lattices to allow for highly sensitive magnetic field measurements.
• Gravimetry based on atom interferometry with new levels of precision in measuring gravitational fields.
• Timekeeping with optical lattice clocks, which have surpassed traditional cesium atomic clocks.
• Gyroscope and accelerometer quantum inertial sensing that provides ultra-precise measurements of rotation and acceleration.
• Rydberg-atom-based quantum electrometry has led to sensors that can measure electric fields with unprecedented sensitivity.
• Quantum-enhanced optical microscopy employs techniques like squeezed light and entangled photons to enhance resolution and sensitivity.
• Developments in quantum-matter exploration and the use of atoms in the development of quantum computers.
In a world where timelines for developing a workable quantum computer remain elusive, there is utility in developing quantum technologies at the edge. More detailed looks at several of these developments and the applications they lead to were reviewed in a recent Forbes event Atom-Based Quantum Technology with Infleqtion.
Convergence
The question is, how is the rich real-time sensing capability coupled with the LLM-based research capability going to be meaningfully exploited? This is where a unified ecosystem of technologies powered by AI can play a significant part, and another recent Forbes quantum computing event, Leading & Innovating with Quantum & AI with SandboxAQ, yielded some strategies for convergence:
• Quantitative modeling can be used today to solve what quantum computation is envisioned to solve tomorrow. SandboxAQ uses LQMs (large quantitative models)—founded on precise equations, approximations and simulations—to model application areas.
• While LLMs are essentially backward-facing based on established research, LQMs are forward-facing in the service of unsolved problems such as the discovery of new drugs.
• Simulations that are causal-based, effectively moving AI from being weak to strong, will prove increasingly important where it isn’t easy to summarize a system’s dynamics via a single equation.
• This LQM-based foundation generates synthetic data that LLM-based AI may be able to leverage to surface insights.
• While quantum computing, via a QPU (quantum processing unit) chip, may be necessary to solve a specific quantum problem (e.g., predicting how two atoms may interact), the QPU will need to be "plugged" into an overarching algorithm written to solve a larger problem.
AI, sitting on top of the technology ecosystem, will be leveraged to manage the overall problem.
A Possible Trajectory
So where will the combination of quantum-based sensing data, LLM-based insight, LQM-based synthetic data and causal modeling, all managed within an AI-enabled environment, lead us?
Here, we can turn to nature for a clue. Evident before us and present in all matter—animate and inanimate—is the atom. A remarkable creation that’s robust, immune to decoherence and scalable, the atom is the quintessential quantum computer. Being so, it fully integrates genetic-type code in its operations. This forerunner—a nature-1.0-type quantum computer—has only captured some of the quantum intelligence that surfaces further in a nature 2.0 type, a.k.a. molecular-based, and a nature 3.0 type (a.k.a. cell-based quantum computers). Hence, the atom and its scaled-up successors are successfully operating syntheses of quantum computation, genetics and intelligence.
This early achievement, vastly prevalent in nature, paradoxically stands as the substance of the boundary toward which technology is inevitably moving. The increasing capability in quantum sensing powered by quantum-matter-type platforms will bring us face to face with the reality that atoms are, in fact, "quantum computing" based on genetic-type code that is driving that computation.
Implications
Powered by continuing advances in magnetometry, gravimetry, optical lattice timekeeping, inertial sensing, electrometry, quantum-enhanced electro-microscopy and the emerging field of atomtronics that will increasingly allow an atom’s wave-like nature to be harnessed, it will become possible to learn to use the atom itself to look even further into the quantum realms, and to so discover new patterns and properties, that LLM- and LQM-type AI will then allow us to make sense of.
This will likely lead to—as suggested by a recent Springer Nature paper—recognition of genres of patterns in life and matter that both fortify and enhance the very structures of physics, chemistry and biology. This will open up myriad additional applications in the material sciences, medical technologies and biohacking realms based on enhancing genetic-type code from the bottom up rather than intervening and making changes from the top down. Further, additional avenues to build different kinds of quantum computers based on insights gained from the direct experimentation and testing of quantum realms, by the plethora of innovative quantum sensors built exclusively to probe those realms, will also open up.
Rather than building superposed and entangled castles in the air, crossing concrete atom-based bridges will lead us to the promise of a quantum AI future.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
{Categories} _Category: Takes{/Categories}
{URL}https://www.forbes.com/councils/forbestechcouncil/2024/10/01/the-future-of-quantum-ai/{/URL}
{Author}Pravir Malik, Forbes Councils Member{/Author}
{Image}https://imageio.forbes.com/specials-images/imageserve/65a96371edadd7947451a2d2/0x0.jpg?format=jpg&height=600&width=1200&fit=bounds{/Image}
{Keywords}Innovation,/innovation,Innovation,/innovation,technology,standard{/Keywords}
{Source}POV{/Source}
{Thumb}{/Thumb}