AI- the New UI?

Manish Choudhary

Evolution of Human-Machine Interaction (HMI)

From digital personal assistants to applications that provide contextual and relevant recommendations for ecommerce, to driverless cars that may soon be a reality, AI is making every interface both simple and smart and setting the bar high for future applications.

The hallmark of Human-Machine Interaction has been marked by the constant endeavor to make machine interactions more ‘human-like’. HMI has come a long way from its ‘command prompt’ days where the interaction was machine-centric to touch screens where users interacted with machines using natural gestures. Today, machines have the capability to understand voice (Natural Language Processing) and sight (Image and Vision Computing) – the two senses which humans use most for communication.

This has been made possible by advances in artificial intelligence (AI) and deep learning that enable machines to interpret, decipher and contextualise human inputs into machine understandable and readable formats.

The next frontier for human-machine and machine-machine interactions

Machines have begun to understand natural human interaction cues with greater accuracy; and interactions have become adaptive and multimodal (type, touch, voice, sight) as compared to the earlier unimodal, passive and command based interactions.

Google Assistant receives nearly 70 per cent requests using natural and conversational language; and around 20 per cent mobile queries on Google are made through voice input. Amazon’s home automation device Echo along with its voice assistant Alexa are breaking barriers by demonstrating the growing adoption of voice-based assistants at home. Word accuracy has also reached the human threshold. Google machine-learning algorithms have increased their word accuracy from less than 80 per cent in 2013 to 95 per cent in 2017, and now match human accuracy levels.

Voice is on the verge of replacing typing. Technologists are cognizant of this fact and are already focusing on building voice-based interfaces in their consumer ecosystems. Led by Apple’s Siri and Microsoft’s Cortana, Samsung is planning to integrate its voice bot ‘Viv’ with all Samsung electronics and household products. Similarly, Facebook is working on Jarvis bot, a voice enabled home assistant that can control home-automation devices.

The new paradigm of interface across the industry

Enterprise applications were the first to adopt AI as the true-north. Not only has the industry seen traction for incorporating self-learning algorithms but has also automated mundane and repetitive tasks by learning from historical data.

AI helps enterprises streamline the workflows and provide services with the best user-experience. Ranging from concierge services (most routine) to augmenting intelligence,

AI is proving to be one of the greatest drivers for increasing workplace efficiency.

Here are some examples from across industries:

• Banking: Banks such as ICICI have begun testing AI chat bots that will act as quasi-bankers and can aid the customers in making their loan choices

• Healthcare: Bengaluru-based AIndra, an AI-based cervical cancer screening start-up is using deep-learning algorithms to interface with pathologists and provide diagnosis based on its machine learning capabilities

• Managing transactions: Taco Bell released ‘TacoBot’ which employs AI and allows customers to order from Taco Bell. It has the ability to recommend menu items, answer questions, organize group orders, and facilitate transactions

• Connectivity: Skype Translator makes conference calls across languages. Neural nets train the service so the ‘voiceovers’ occur fast enough to support a live conversation

• Intelligence Augmentation: Salesforce Einstein acts as a personal data scientist as it learns from existing data, and delivers predictions and recommendations based on the unique business processes

Text-based chats, spoken conversations, gestures, or virtual reality are varied ways through which AI connects with consumers. AI is rapidly changing the way we interact with our surroundings and is becoming omnipresent. From cashier-free, self-checkout grocery shopping to security checks at the airport, AI is enhancing user experience across the spectrum.

Today, making machine interaction indistinguishable from humans is not a chapter out of science fiction, it’s imperative for anyone looking to stay in business.

(Manish Choudhary is SVP, Global Innovation and MD, India Operations, Pitney Bowes Inc. Views expressed here are personal)