The Future of AI is Emotional

Jose Ferreira, SVP, Product & Innovation | Adam Ghahramani, Associate Director, Decision Sciences

October 13, 2021

bottom curve

We have transcended the age of “Big Data” where collection and scale were primary competitive considerations, to a new paradigm where strategic applications of data, mostly in automating and scaling processes that were once done manually. Many of the more interesting trends in this new frontier of data engineering are happening within the realm of artificial intelligence (AI). AI has already started to disrupt many industries, including marketing and media, by enabling more personalized customer experiences across traditional and leading-edge media channels.

We are all exposed to AI daily, often without even realizing it, as AI-powered technology has become ubiquitous and seamless in our everyday lives. A staggering majority of companies use AI in some way; it has value in almost every function, business, and industry. Specific to marketing, AI technology is used to track consumer behaviors and use the data to inform more personalized and relevant content and channel executions. It is extremely common to get to the bottom of a website and see “If you liked this, you might also like…” recommendations. All social networks and large streaming video platforms rely heavily on this application of AI to keep audiences engaged, interested, and in platform. In the fitness sector, Under Armour uses data from its Record app to deliver more personalized training to their users. In the travel industry, AirBnB uses a machine learning algorithms that optimizes pricing in real-time. It was able to do this by looking at a buildings features, location, and other traits while comparing it to nearby places. AI also promises to disrupt the healthcare sector in a multitude of ways.

AI in Healthcare

Insider Intelligence reported that spending on AI in medicine is projected to grow at an annualized 48% between 2017 and 2023. AI is expected to improve healthcare by fostering preventative medicine and new drug discoveries. Two examples of how AI is impacting healthcare include IBM Watson’s ability to pinpoint treatments for cancer patients, and Google Cloud’s Healthcare app that makes it easier for health organizations to collect, store, and access data. According to Insider Intelligence, 30% of healthcare costs are associated with administrative tasks. AI can automate some of these tasks, such as pre-authorizing insurance, following-up on unpaid bills, and maintaining records, with the goal of decreasing the workload of healthcare professionals and ultimately saving the company money.

AI solutions are also used in medical imaging, aiding cardiologists and radiologists by surfacing relevant insights from tests. This helps the medical professionals identify critical cases first, make more accurate diagnoses, and potentially avoid any errors. Similarly, the electronic healthcare mandate of the Affordable Care Act has allowed AI to quickly assess millions of health records and potentially expose healthcare conditions or issues that a human wouldn’t be able to glean without the computational power of an AI. These innovations don’t come without challenges. Adopting AI into the healthcare space requires meeting regulatory requirements, as well human trust in the accuracy of results , and developing a willingness for humans to cede some control of processes that are better served by an AI. These examples, while successful and disruptive, are leading to more sophisticated applications of AI – the kind of stuff prophesized in science-fiction movies, a more human AI or Emotional AI as it’s commonly described.

Emotional AI

Video has grown to be technologically ubiquitous and socially omnipresent. Anyone who was working in a business office setting pre-pandemic is now interacting almost exclusively with colleagues through video conferencing applications. These advances have allowed us to see and hear each other with the click of a button but are just starting to tackle the larger problem of understanding non-verbal communication. In fact, 90% of all communication is non-verbal—cues, gestures, and reactions to speech that are sometimes hard to understand. This is where Emotional AI modeling is being trained to fill this void. The Chief Science Officer of Affectiva, Dr. Rana el Kaliouby, recently spoke at SXSW about the many advances her organization is making to re-imagine human-to-human interaction. They are in essence objectively scoring human emotions by way of non-verbal cues to enhance conversation. In practice this could mean a video conference between colleagues that is enhanced by a feed of emotional response text. Or in a personal setting two people could wear Google Glass-style devices that give them insights into emotions the other is not verbalizing. Companies like Soul Machines are already operationalizing some of these concepts. They build on demand hyper-real ‘digital people’ that can be used by brands to have near-human digital and video interactions with live customers, thereby improving the overall experience. It takes the idea of the site embedded chat bot to a new level of AI-powered personal interaction that includes non-verbal elements. It all sounds a bit scary when you game out the nefarious ways the technology could be applied, but used within the bounds of a coherent regulatory framework of willful participation, it holds a lot of promise for the healthcare sector.

Imagine a virtual or live consultation with a healthcare provider where a patient may be worried or embarrassed about a specific topic, which the healthcare provider can pick up on through non-verbal signals and proactively probe to provide better care. In a mental health practice setting it could significantly enhance a practitioner’s ability to diagnose patients and provide short- and long-term care. People with a condition that hinders their ability to communicate could see marked improvement in quality of life for themselves and peer groups. And even for those who are able to fully communicate, a wearable that measures data related to physical or mental health may be able to show a clearer picture via data than what a typical patient is usually able to convey.

On the commercial side of the healthcare business the technology is already in use, in discrete cases, to test emotional responses to different stimuli. That allows marketers to think beyond trackable behaviors to more important motivational and emotional metrics of success. AI has been used in healthcare settings to optimize efficiency of healthcare delivery, and this next generation of emotion driven AI holds the potential to significantly disrupt and improve both health outcomes and personalized communication strategies. We continue to closely monitor developments in this exciting space and build out novel use cases of the underlying technology.