IonQ CEO sees quantum computing's application in machine learning taking the lead – Washington Business Journal – Washington Business Journal

Every conversation around quantum computing and its potential to revolutionize industries eventually circles back to two questions: When will it outpace classical computers and where will it be applicable first?
For IonQ Inc. President and CEO Peter Chapman, the former question’s answer is complicated. But he thinks he may have an answer about the latter. 
That’s because IonQ (NYSE: IONQ) is seeing promise in quantum computing’s application to machine learning (ML) and how it can outpace current, often referred to as “classical,” computers, Chapman said Wednesday at the University of Maryland’s first Quantum Investment Summit.
“We put in our investor deck that we thought by 2023 that machine learning would start to be the first kind of thing to fall for quantum computers. But just recently, in these last nine months our so, we’ve been doing quite a bit of machine learning,” said Chapman, whose College Park company just began trading Oct. 1 on the public markets. “We’ve now seen in three areas where we can beat the best classical result in machine learning, even though the quantum computers themselves are not more powerful than the best classical systems.”  
Machine learning, effectively seen as a subset of artificial intelligence, uses algorithms to analyze various data sets and model possible predictive outcomes, such as when an email account offers autocomplete features to predict the next word in a message that a user is composing. 
Related: Read our cover story on the D.C. region’s growing quantum computing industry — and why it matters
Such ML automation tools have been in hot demand, both in the private and public sectors. But due to the prohibitive expense of training machine-learning models, companies aren’t able to update them fast enough to incorporate new developments, like the Covid pandemic, into their algorithmic predictions, Chapman said.
But with the potential gains made by quantum computers, the cost of training a machine-learning model drops and its possible returns grow.
“In the classical, [graphics processing unit]-based system, it took 20,000 iterations to create the model. But with a quantum version of it, we did it with only 26 iterations,” Chapman said. “If you can train the model in a thousand times fewer iterations, that means a bunch of things. One, that means we can create bigger machine learning models. Two, it looks like we can create better ones, and three, the cost of those things can come down.” 
Chapman went on to note that with quantum-powered machine learning, companies across various industries and sizes could capitalize on the automated technology, instead of just the largest firms like Amazon.com Inc. (NASDAQ: AMZN) or Alphabet Inc. (NASDAQ: GOOG).
To that end, IonQ has already struck a partnership with Fidelity Center for Applied Technology (FCAT) — Fidelity Investments Inc.’s technology research & development lab — to author research on how quantum computers can outperform classical ones in producing high-quality data for testing financial models
The company has also done research with General Electric Co.’s (NYSE: GE) GE Research exploring quantum applications for risk analysis and the Goldman Sachs Group Inc. (NYSE: GS) on how to apply the technology to Monte Carlo simulations, or computerized models used to predict various outcomes across such areas as finance or supply chains. 
Chapman said the application doesn’t represent the oft-discussed barrier of quantum advantage — the theoretical moment when quantum computers can solve mathematical problems that classical ones can’t — but it does present an opportunity for quantum to augment the capabilities currently in the market. 
“The customer, at the end of the day, doesn’t care about quantum supremacy or quantum advantage or anything else,” he said. “All they know is they have a trading algorithm, they’ve been working on that trading algorithm for 20 years to get it to be the best that it can, and some young pups show up with a quantum algorithm and manage six months later to beat that algorithm.”
© 2021 American City Business Journals. All rights reserved. Use of and/or registration on any portion of this site constitutes acceptance of our User Agreement (updated 1/1/21) and Privacy Policy and Cookie Statement (updated 7/20/21). The material on this site may not be reproduced, distributed, transmitted, cached or otherwise used, except with the prior written permission of American City Business Journals.

source
Connect with Chris Hood, a digital strategist that can help you with AI.

Leave a Reply

Your email address will not be published. Required fields are marked *

© 2021 AI Caosuo - Proudly powered by theme Octo