Founder and Principal Analyst, ZK Research with a focus on emerging technologies that enable organizations to transform digitally.
I’ve always been a big advocate of artificial intelligence (AI) and its near-limitless potential to improve our lives. For example, society has already come to enjoy many benefits of audio AI technology. When I was in college, students had to choose between paying attention and taking notes in class, because it was too difficult to do both. Some students would record lectures, but then they’d have to waste time manually transcribing them later.
Advancements in AI audio transcription have not only democratized fast, affordable transcription for students and professionals; they’re driving globalization for some of the world’s biggest and most influential organizations. Now, some companies even offer real-time translation capabilities that support multilingual virtual meetings, breaking down language barriers at scale.
Today, data AI is pretty good and voice AI is getting a lot better. But we’re just starting to scratch the surface with video AI. Not only does video AI technology have the potential to recoup the benefits of being together in pre-pandemic offices and classrooms; it’s creating virtual work and learning environments that are even more engaging than the physical spaces we’re leaving behind.
Tapping Enormous Potential
I’ve already experienced the benefits of basic video chat firsthand. For example, when I used to have 10-person businesses meetings in an actual conference room, I had to physically arrange business cards in front of people so that I could remember their names. Now, with virtual meeting platforms like Zoom, I only have to look at the screen to see someone’s name associated with their face.
MORE FOR YOU
There’s much more on the horizon. Cisco recently announced its mission to make virtual communication 10 times better than being in person. The latest version of Webex includes AI-powered gesturing capabilities, allowing participants to send reaction emojis (like a smile or thumbs-up) by emoting the way they usually do instead of having to hit a button. Last month, conversational service automation (CSA) company Uniphore acquired Emotion Research Lab, which offers sentiment and engagement analytics that can measure people’s engagement and signal emotional changes.
Compelling Early Use Cases
These early video AI technologies have an array of compelling use cases in education, medicine, sales and customer service. AI tools can alert teachers when students aren’t paying attention and improve doctor-patient communication regarding sensitive test results and diagnoses. AI can offer real-time insight into customer engagement and suggest ways to improve customer experience, which is the number one brand differentiator today. Not only can AI improve sales communications; innovation such as NVIDIA’s DRIVE IX will also power new life-saving products like AI-assisted cars that make driving safer.
In order for video AI technologies to take hold, they must be intuitive to the user. The point of video AI is to transcend technical know-how and make it easier for the user to communicate as well as receive feedback about themselves and others. AI has the potential to help all of us improve our communication by showing the real-time effect our verbal and nonverbal communication has on others.
In this vein, video AI has therapeutic potential for helping people with social and developmental disorders learn to communicate more effectively. For example, AI technology could assist children with Autism Spectrum Disorders in understanding classroom social cues.
Barrier To Entry
There will, of course, be pushback to video AI by people who distrust its accuracy. AI certainly does make mistakes, but the barrier to entry for video AI isn’t perfection; AI only has to be more accurate than humans in order to be useful. In fact, implementing these technologies is the best way to improve their accuracy. AI works by learning algorithms; the more data it’s fed, the bigger its learning set and the higher its degree of accuracy.
If you’re a CEO using sentiment and engagement analytics to give a virtual speech to 3,000 people, will AI interpret the facial expressions of every person in the audience correctly? Probably not. But even if it’s only 97% accurate, it will correctly gauge the speech’s emotional effect on over 2,900 people, which is exponentially more data than the human eye can observe.
Even AI with a moderate degree of error can provide better feedback than humans. From a business leader’s perspective, it’s important to understand that artificial intelligence will make mistakes, but once it’s more accurate than people, it becomes a vital tool that could not only reshape businesses and entire industries, but substantially improve how we live in the new post-pandemic virtual world.