Pages Menu
Categories Menu

Posted in Uncategorized

The Duality of Machine Learning

By Judy Davies, Vice President of Global Marketing Communications, Advantest America

The term “binary,” with which we in the semiconductor industry are quite familiar, refers to more than the 1s and 0s found in binary code. It implies a balance, a duality that is present throughout the industry. This duality is found in our human makeup, as well. We use both intellect and feeling in living our lives, as we identify challenges and determine solutions.

If artificial intelligence and machine-learning systems are to truly think as humans do, it would seem that moving beyond purely digital computations will be essential. This means finding a way to teach machines to combine left-brained (analytical, data-based) with right-brained (intuitive, perception-based) thinking – i.e., the true duality of the human brain.

The work of John von Neumann has come to represent the left-brained approach. Beginning in the 1920s, von Neumann applied his genius in mathematics across a wide spectrum of projects. These included working on the Manhattan Project to construct the first atomic bomb; creating the landmark von Neumann architecture for digital computers that store both programs and data; and developing the field of game theory, which many high-stakes poker players use today to deduce future outcomes and win tens of millions of dollars.

The right-brained approach can also be described as emotional intellect. It represents more analog or interpretive thinking that takes into account human feelings and attempts to inform actions that are difficult to quantify. As an example, whereas von Neumann’s game theory is used to arrive at decisions through logical reasoning, poker players also gather information about their opponents by reading their body language and demeanor at the table. This is the right brain at work.

Neuromorphic computing involves making machines that more closely replicate the way the way the human brain works. Rather than being limited to solely digital processing, neuromorphic chips assimilate analog information, which is then interpreted for shades of meaning. This forges a path to creating neural networks that are aligned with how we think.

Already present in our lives is what can be viewed as a precursor to neuromorphic computing. When we visit an online retailer’s site, our interest in the products viewed and/or purchased is catalogued, grouped with the interests of other buyers, compared with those buyers’ previous purchases, and used to pitch us on buying other products that people within that demographic have bought. Pop-up ads, emails and texts claiming “You may also be interested in …” demonstrate how computing power is being applied to get into consumers’ heads and not just understand but influence their spending patterns.

Similarly, machine learning can be applied when it comes to guiding consumers’ future actions. Databases are being used both to predict our needs and to stock local inventories accordingly, ensuring that our local store or distributor will know as soon as we exhaust our supply of a particular item and will be able to offer same-day delivery of a replacement.

Factoring in product reviews from other members of our demographic group would allow retailers to draw high-probability conclusions about both our level of satisfaction with products we’re currently using and the likelihood that we may be willing to switch to a similar product from a different supplier. This educated guesswork will be based on “reading” your emotional decision-making processes. With this ability to predict future behavior, poker-player computers are assured continued dominance.

The state of the art in neuromorphic computing does not yet involve precisely predicting all of our next moves. The world of the Steven Spielberg movie “Minority Report” – in which savant-like “pre-cogs” can predict future crimes before they occur, enabling law enforcement to arrest criminals-to-be in advance – does not yet exist. But it’s intriguing to consider, and to wonder if we may actually get there at some point.

Would bringing the duality of digital processing and emotional intellect to fruition be highly beneficial, enabling digital assistants like Alexa and Siri to more accurately anticipate our desires? Or would it bring us a step closer to having our lives actually be run by the machines in our lives? One thing seems sure: If and when full-blown neuromorphic computing becomes a reality, it will definitely be put to use.