How quantum computing and smart planning could supercharge AI

The phrase “quantum computing” sounds like something from a science fiction show, but it’s very real. Some federal agencies are already investing in it and more are sure to follow. Agency leaders need to understand what it is, how it differs from the traditional computer technology everyone uses, and what to do in order to implement it.

Put as simply as possible, quantum computing uses quantum mechanics to perform computation. These phenomena enable certain kinds of calculations at a speed and scale that conventional computers cannot come close to matching. However, much of its potential remains theoretical as scientists continue to refine the technology.

In particular, quantum computers can perform integer factorization quickly—that is, finding two previously undetermined prime numbers that produce a known number when multiplied. It might sound like an abstract math puzzle, but it’s actually an important aspect of encryption/decryption and artificial intelligence (AI). (A more complete explanation is beyond the scope of this post.)

President Trump issued an executive order in February 2019 titled “Executive Order on Maintaining American Leadership in Artificial Intelligence,” defining AI as a priority in research and development. While it does not specify quantum computing, it does provide a strong motivation for government agencies to explore the technology as a means to accomplish the end.

Jason Porter took part in a panel discussion on AI’s role in fighting fraud, waste and abuse on Government Matters TV on WJLA on June 30. View the show on demand here.

Quantum computing, and related technologies such as hyperconverged infrastructure and scalable computing, provide ways to deliver powerful AI; but before bringing quantum into the picture, it is important to think through your use case and determine what you really need.

Define the outcomes. What are you looking for the AI to tell you? The goal of AI is to get the computer to think a bit on its own. To do that, you have to be specific about your desired outcomes. If you do not clearly define the desired outcomes, AI and machine learning components may provide skewed or misleading information.

Build for what you need. AI and ML are similar, but they are not the same.