This article was originally published on Influencers
Artificial intelligence and machine learning promise to fundamentally change how businesses deliver services to their customers. Greater personalization, customization, and predictive ability are all part of this promise, which is leading executives to dramatically increase spending on their AI capabilities resulting in market analysts predicting an AI software market worth $37 billion by 2025. Other forecasts predict that revenues for AI services, software and hardware will grow 16.4% year-over-year in 2021.
However, here at Globant, we believe that simply investing in AI skills and capabilities is not sufficient. The increase in customer expectations has risen to such an extent that people now expect what we’re calling “human-centered AI”. For too long, businesses have focused on using technology to create more efficient products and services, automating everything possible, with the objective of providing fast and efficient services but have ignored the human element.
Human-centered AI experiences are becoming important due to two opposing forces:
- Research shows that people respond to and treat technology as if they are interacting with another human being. This is based on the media equation theory, where for example people describe robots in human terms, or provide polite answers to computers. People want technology to help them in their everyday lives, and when it does, quickly becomes an integral part of them.
- However, 59% of consumers feel that companies have lost the human factor in the customer experience. This is generally due to poor or ineffective use of technology. For example, while speech recognition and language processing technology is highly advanced today, we’ve all had frustrating experiences interacting with a chatbot, virtual assistant, or voice recommendation system. By not designing a conversational flow similar to the interaction between two people, properly handling possible errors of interpretation and understanding, the human factor is lost and the experience is impoverished. That happens with many AI-based systems.
More human AI-powered experiences mean learning to blend different elements and technologies. For example, rapid advances in natural language processing (NLP) mean that it’s now possible to provide people with highly customized answers in real-time, for example when interacting with a bot or other conversational interface.
Human-centric AI means getting away from the use of the standardized applications that we have become so used to. A bank’s application looks the same to you, as it does to your retired parents, or teenage children, despite very different financial needs and objectives. Similarly, a retailer’s application will often have the same bland look and feel, as it is used by all the different demographics that use the retailer. However, with the use of AI we can customize the user interfaces of these applications depending on the characteristics of the client or the use that will be given to the application.
Human-centric AI also means understanding the so-called “black box” of AI, that is what happens behind the scenes, and how the algorithm is reaching decisions. In order to trust AI-based systems to make more and more complex, yet vital, decisions it’s essential that all stakeholders understand at least at a high level how the system is working. This will be important for everyone from business executives using the inputs to make strategic decisions, to regulators who will need to understand inputs in order to trust results. We’re seeing approaches, such as LIME (local interpretable model-agnostic explanations), which provide ways to better understand your machine learning models.
There is no doubt that experiences based on human-centric AI will be essential to build an emotional bond between brands and their customers. They will also play a key role in the wider adoption of AI-powered products and services. Achieving this means keeping people in the game, focusing on creating a “customer journey” that creates an intimate connection between people and AI.