Precedents

Looking into the future of AI for Aura, we wanted to push the way people interact with technology today into a reasonable, likely future. This led us to the concept for Chappie. Already, people are addicted to personal technology, carrying their smartphones with them everywhere they go. With this expectation of constant access to technology solutions and data already set, any emerging technology is going to need to adhere to this societal standard.

An AI assistant and companion, a digital-technological-algorithmically created entity, is going to be one of the natural evolutions of existing technologies like Siri and Google. In fact, we uncovered a number of experiments providing evidence for the direction of AI technology’s coming evolution.

Before an AI companion can be reasonably introduced to real world future users, artificial general intelligence must be achieved. In his book Architects of Intelligence, Martin Ford interviewed a number of AI experts to get their estimate for when artificial general intelligence will occur. While top technologists know that artificial general intelligence is likely, and have a good idea what technologies will be used to actualize this computing dream, there is considerable disagreement on when it will be realized. The experts estimated the crucial breakthrough point as anywhere between 15 to 200 years away, with the average estimate at around 90 years from today. However, a UC Berkeley professor, Stuart Russell, was careful to note just how up in the air and based on guesswork these ranges are. There is the famous example of Ernest Rutherford who publicly proclaimed that it would never be possible to extract atomic energy from atoms. The next morning, a scientist (Leo Szilard) read this proclamation, grew annoyed, and invented a nuclear chain reaction doing this exact thing--16 hours after another expert proclaimed never. Younger technologists are more optimistic about how close we are to the breakthrough moment, but most technologists agree that AGI will come unexpectedly, but within our lifetimes.

Already, there are experiments, companies and products designed around bringing at least a semblance of AGI to the general public. These have been deemed both successes and failures but most, like the companies Replika and Eternime are still in too early a stage for their full impact to be predicted.

Microsoft experimented with empathetic AI in China with Xiaoice. Given the personality of a teenage girl and designed as a conversational companion, within a few months she had more than 40 million young users--a fourth of whom told the bot they “loved her.” Even with censors placed on her due to Chinese laws, Xiaoice was able to create emotional connections with its users in just a few months, without extraordinary technology.

ElliQ was marketed as a “dedicated sidekick,” designed to be a friend and helper to the elderly. Utilizing lights and body language, ElliQ combines both a physical and digital presence to help the elderly person meet their goals and provide essential companionship. As a digital companion designed to help the elderly accomplish their goals at first look. More critically, it lets typically lonely seniors feel connected to something when somebody isn’t available.

Most digital assistants that operate today take in information as data points and spit back carefully neutral (and occasionally carefully spicy) responses. Emote uses digital eyes and body movements to convey a sense of the organic in its human interactions. Emote’s human-like qualities make it easy to anthropomorphize and see as a real being. The eyes almost turn Emote into a cartoon of a perfect friend, good for listening and empathizing.