The promise of quantum computing revolutionizing artificial intelligence, particularly deep learning, has been a topic of great interest in recent years. However, a growing consensus suggests that this may not be within reach as quickly as expected.
According to Filippo Vicentini, Assistant Professor of AI and Quantum Physics at Ecole Polytechnique, the difficulties in processing information from neural networks and voluminous data hinder the advancement of AI through quantum computing.
Despite companies like Google, Amazon, and Microsoft promising to solve small real-world problems within the next few years, quantum computers’ limitations, including their slow speed and short calculation capabilities, make them less compatible with AI than initially thought. Researchers like Torsten Hoefler, Thomas Häner, and Matthias Troyer highlighted these challenges in a recent paper. Meanwhile, machine learning is becoming an essential tool for designing and operating quantum computers, with companies like Q-CTRL pioneering this field.
The Promise of Quantum Computing for Artificial Intelligence: A Reality Check
The idea that quantum computing could revolutionize artificial intelligence, particularly deep learning, has been gaining traction in recent years. However, a growing consensus suggests that this may not be within reach anytime soon. Filippo Vicentini, Assistant Professor of AI and Quantum Physics at Ecole Polytechnique (IP Paris), explains that the difficulties in processing information from neural networks and voluminous data make it challenging for quantum computing to advance AI.
One of the main reasons behind the belief that quantum computing could boost AI development is the similarity between how neural networks are trained in deep learning and how early quantum computers were tasked with doing many short “quantum” subroutines. The hope was that a reasonably sized “quantum circuit” would be more expressive, meaning it could present more complex solutions to a problem with fewer resources, thanks to quantum phenomena like interference and superposition.
However, experts are recognizing that quantum computers will likely remain very slow when it comes to data input and output. Even if we are optimistic, a quantum computer that could exist maybe five years from now would have the same speed to read and write as an average computer from 1999 or 2000. Moreover, the output of a quantum computer is probabilistic, which creates additional challenges.
The Limitations of Quantum Computing for Big Data and Neural Networks
The consensus is growing that using quantum computing for big data and neural networks may ultimately not be worth the effort. A recent paper by Swiss National Supercomputing Centre’s Torsten Hoefler, Amazon’s Thomas Häner, and Microsoft’s Matthias Troyer laid out this position. The tone of the quantum machine learning community has been on a downward trend.
One of the main limitations is that quantum computers will likely remain very slow when it comes to input and output of data. If we try to run quantum computers faster to increase the amount of data we can inject, we will start to introduce more errors in the calculation, and the result will deteriorate. There seems to be a speed limit for the operation of these machines above which the noise and errors are too strong to be corrected, even when we look about 20 years in the future.
The Role of Machine Learning in Quantum Computing
Despite the limitations of quantum computing for AI, machine learning is quickly becoming an essential tool to learn how to design and operate quantum computers nowadays. For example, every device is slightly different, and reinforcement learning techniques can analyze your machine and its particular patterns to help fit algorithms specifically to that device.
Companies like Q-CTRL, Google’s Quantum AI, and Amazon’s Braket are leveraging these ideas. Moreover, AI could also be very complementary to quantum computing. Microsoft’s Azure Quantum Elements used a combination of Microsoft Azure HPC (high-performance computing) and AI property-prediction filters to whittle down a selection of 32 million candidates for a more efficient rechargeable battery material to just 18 candidates.
The Future of AI and Quantum Computing: Complementary but Not Compatible
In conclusion, AI and quantum computing will be different components in a stack of complementary but incompatible tools. While quantum computing may not revolutionize AI as previously thought, machine learning is becoming an essential tool for designing and operating quantum computers. The future of AI and quantum computing lies in creating joint teams that can leverage the strengths of both fields to push the boundaries of what is possible.
As Filippo Vicentini puts it, “We want to keep pushing those directions and many more by creating a joint team called ‘PhiQus’ between Ecole Polytechnique (IP Paris) and Inria together with Marc-Olivier Renou and Titouan Carette.” The future of AI and quantum computing is bright, but it will require a nuanced understanding of the strengths and limitations of each field.
External Link: Click Here For More
