Huawei Consumer Business Group has now revealed a study, conducted for the company by Lightspeed Research in November as part of the launch campaign for its Kirin 970 A.I. chip, which Huawei claims to show the similarities between how human and artificial intelligence work. The results shouldn't be altogether surprising, with consideration for prior studies conducted into conscious decision-making as compared to subconscious decision-making over the past decade. With that said, although the similarities between how human brains and A.I. processing function are debatable, the study could provide some insight into the reason some current A.I. implementations have been successful. Furthermore, it could have implications for the possible uses of A.I. moving forward.
This new study was conducted as a poll, which took input from 10,000 individuals across Europe, in the UK, Italy, Germany, Portugal, Spain, Netherlands, Ireland, Switzerland, Belgium, and France. Specifically, Lightspeed Research determined that the average European is not fully aware of as many as 99.74-percent of the decisions they make. Rather than the majority of decisions being made consciously, the bulk of processing and choosing actually appears to happen in the background. That conclusion is based on the commonly accepted observation that the human brain is actually responsible for around 35,000 decisions every day. Respondents, on the other hand, were only able to account for around 92 decisions per day. While a research poll is certainly not the most scientific approach to making discoveries, the new findings do appear to track closely with previous scientific endeavors.
To begin with, the average respondent guessed that they made around seven decisions related to food on a daily basis, while the actual number of decisions made related to food is said to be around 221. Interestingly, that not only further highlights the unconscious nature of decision-making. It could also be correlated to the fact that, when asked, 47-percent of respondents admitted that they would like a smartphone-based system to help them discover new ways to use the food that is in their fridge. What that may reveal is that while people do not, in fact, put very much conscious thought into food, they would likely welcome an A.I. that augments the decision-making process as it pertains to food.
Respondents also revealed a disconnect in other areas of conscious thought that could be tied to an explanation of the rapid growth and advancements in A.I. According to the poll, respondents estimated that they check their smartphones an average of 22 times per day. That falls well short of Huawei's claimed average, which is 76. Meanwhile, 43-percent of respondents also responded that they would like automatic notifications about travel, which is a function already served by some algorithmically-driven digital assistants. Beyond that, 39-percent of respondents admitted a desire for assistance in learning new languages, 38-percent for taking better photographs, and 31-percent for being introduced to new music based on their established tastes.
Each of those is, perhaps unsurprisingly, an area where technology and A.I. solutions are already being applied or beginning to be applied. Moreover, they don't fall too far outside of Huawei's own goals with its above-mentioned Kirin 970. The chip's primary purpose is to enable smartphones to maintain optimal performance for longer through A.I.-driven settings optimizations and device regulation. However, according to Huawei executive Walter Ji, A.I. is also intended to make lives easier and to enable users to make "better, more informed decisions." Huawei's sentiments about the importance and potential of A.I. are shared by Paul Dolan, a professor of Behavioural Science at the London School of Economics and Political Science. Commenting on the study, Dolan says that the results serve to demonstrate the fact that human minds often act unconsciously, with habits being formed to prevent overload from being caused by everyday tasks. He goes on to posit that the results show how A.I. actually mimics that subconscious activity. That, he says, lends to its potential with regard to improving the decision making process and enabling a higher degree of happiness.
It is, of course, important to note that the underlying differences between how a computer and a human think remain. Human thinking is primarily driven by emotion, environment, and bias, while A.I. is computational and must be programmed. Setting aside the ethical debates surrounding A.I., however, there are also likely to be a large number of other areas where augmentation of human thinking would or could be useful that simply haven't been considered yet. With that said, respondents to the study do appear more than ready to accept additional assistance in those areas where A.I. could be implemented.