AI breakthrough: The Disruptive Technology in 2019

One of the hottest buzzwords of the current era is “Artificial Intelligence (AI)”, and may we have never seen such a rapid technological change that we are currently witnessing. Of course, much more yet to come, Gartner Predicts 70 Percent of Organizations Will Integrate AI to Assist Employees’ Productivity by 2021. And according to PwC by 2030 AI could bump up the global GDP by $15.7 trillion or a 14% increase. Also, AI would raise employment by 10 per cent.

Already AI techniques are helping us to determine search results, translates our voices into meaningful instructions for smartphones and PCs and more. In the next few years, we can expect to see AI driving our cars answering our calls and many other activities.

In the year 2019, we can see all of this and more but what could be the major AI breakthrough, let find out…

Dueling AI

It's a simple idea: Want to make your future AI smarter? Have it battle it out with another AI. In this case, the arena is digital images: One AI attempts to create a realistic image, and another AI attempts to decide whether the image is real or artificial.

Of course, this is just one domain in which AIs can "duel." Any domain can be modeled by computers: voices, video or whatever you might want to work with. The concept is called "generative adversarial networks" (GANs).

Imagine how that will advance online verification, such as today's CAPTCHA technology, which requires real people to identify objects in grainy photos. With enough dueling, systems can sharpen their wits enough to easily break that sort of gatekeeper software. That will mean a new reality for secure browsing online (another reason why blockchain may be necessary).

That's just the tip of the iceberg. AIs are intended to be intelligent systems that can, in principle, be unleashed on any problem domain. Rapid and independent "self-improvement" of AIs through dueling may lead to breakthroughs in medicine, technology, transportation or other important areas of life. In fact, tech giants like Amazon and Alibaba are already diving into research in this area. Arush Sogani, Director, Sysnet Global Technologies shared his thoughts.

Rajan Navani, Managing Director and CEO, JetSynthesys shares his thought on dueling - The greatest breakthrough in AI that can disrupt the tech world in 2019 is probably Dueling. It’s a simple idea that tackles one of the greatest challenges faced by data scientists when it comes to AI, how do you make a machine imagine and create something new? While dueling, one neural network is forced to create images as the other differentiates between real or fake, pushing it to imagine. The next well be training AI through smaller datasets, currently AI requires a large dataset in order to learn which, while in use as the user feeds it data, takes time to gather. The above two combined, teaching machines how to be innately creative and teaching them to learn largely independently will be the real-game changers, especially for the digital industry.

The focus of AI will shift from cold intelligence to empathy

According to Suman Reddy Eadunuri, MD, Pegasystems India, “We’re moving beyond the point where cold analytic intelligence suffices for consumer-facing AI. Customers want to know that they are being viewed as individuals and not just as customer data records. In 2019, vendors will focus more on increasingly humanizing AI with emotional intelligence and empathy."

This means better understanding context and longitudinal conversations, picking up on clues on customer motivation and intent, the tone of voice, how they feel in the moment, where they are in their life stage, how they act in certain situations, and even what is happening around them. It also means effectively fusing AI systems with human agents, so the wisdom of AI and human crowds are used to deliver intelligent and empathetic customer experiences.

Understand videos as humans do

Aravind Tambad, Head of Growth & BD, Charmboard shares his thought for the video content industry that “the new breakthrough is when the AI’s prowess in making videos shoppable is done at scale and efficiently. The question is can machines understand videos like we humans do, and if so, can the same be done at scale and without errors? Achieving the same breakthrough will make the solutions commercially viable and accessible to all and we already have the solutions to enable the same.”