As we move from a world of smart devices to a world of intelligent objects and services, there seems to be an increasing misconception – or overinflated expectation – about what this intelligence means and how it can help consumers. According to Ovum's latest Digital Consumer Insights survey, 55% of US internet users use digital assistants on a regular basis, which suggests a very positive outlook for consumer-facing artificial intelligence. However, a closer look indicates that many people do not use digital assistants for their more intelligent functions: the two most popular daily use cases for digital assistants are to play music and switch the TV on and off.
Figure 1: How often do you use the following functions of your digital assistant?
A clear distinction needs to be made between what constitutes a new user interface (i.e. voice commands) and what intelligent tasks digital assistants can perform, such as making a booking on your behalf, navigating to a location, or finding a picture in your family album. These use cases exist today but aren't popular. Why is that? The ability to perform intelligent tasks is insufficient in itself to convince people to use digital assistants; there needs to be clearer practical benefits to these abilities.
Amazon has put tremendous effort into fostering innovation in this space, with more than 50,000 Alexa "skills" developed so far. But none have seen a high level of usage besides voice-capable implementations of popular radio and music apps (e.g. BBC, Spotify) or jokes and games (e.g. 4AFart, Meow!, Guess My Name). Instead, practical AI should work in sync with consumers to help them perform daily tasks better – a watchful eye over your shoulder to help you manage your busy digital life. Examples of this are starting to emerge via email clients from Microsoft and Google, which notify you if you forget to include an attachment or remind you to reply to an email (see Cortana's Heads-up feature).
Another example of practical AI is the camera app on the iPhone XS which automatically detects portrait scenes and adjusts camera settings to take beautiful photos. Image recognition is also used by Google Photos to automatically organize your entire photo library and beautify some of your favorite shots. These are valued practical uses because they help with productivity or embellish aspects of everyday life. The value that people attach to consumer-facing AI is very important because it will determine future monetization opportunities, or lack thereof.
Simply put, "intelligence" is the ability to observe and learn rather than interpret and mimic. Consumer AI stakeholders need to rethink their design strategy and develop products and systems that correct our mistakes, help us memorize, help organize our digital life, and help us to make important decisions – like a wise old friend or relative. Truly intelligent technology doesn't need to be clever, just helpful.
Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please contact us.