"AI Map: From Everyone, for Everyone" --- A report on The survey “Hello! AI World - What do you want to do with artificial intelligence? -” was conducted from February 3 to 24, 2020.
AI Map: From Everyone, for Everyone
Miraikan collected Japanese citizens’ various opinions on the perception of AI-related technologies though the survey “Hello! AI World - What do you want to do with artificial intelligence? -” with our survey system “Opinion Bank” from February 3 to 24, 2020. Then we visualized Japanese citizens’ various frank opinions and perspectives as “AI Map: From Everyone, for Everyone”.
What is “AI Map: From Everyone, for Everyone”?
“AI Map” visualizes trends of opinions and perspectives, and hopes and fears on AI-related technologies on the basis of frank voices from Japanese respondents. Let’s look at this map and compare your own opinions with those of others. It will be a clue to think about the relationship with AI, which will sink more to our life in the future.
In this page, we will introduce you about this project and how to view this informative AI Map.
What were the most popular categories in this survey?
There are 9 life events or life stages that may be associated with AI technologies: "Love", "Job selection", "Election", "Marriage", "Child-rearing", "Medical care", "Disaster prevention", "Nursing care" and "After death". Respondents were asked to select three categories (themes) from them. Tabulating and showing the results on a graph, you may find the more popular categories.
The most popular was “Medical care”. Around 76% of respondents chose it. The second-most popular category was “Disaster prevention”, and the third was “Nursing care”.
These three categories are relevant to issues for Japan’s social system, implying that many people expect the application of AI to social issues. This tendency is irrespective of age; even young people were apt to select these categories.
More feedback on AI as seen on the survey
Analysis of the survey indicated the following features.
- People in 10-29 age group showed a trend of disagreement to use AI more than older age groups.
- In the category “After death”, asking whether AI should be used to faithfully recreate the words and behaviors of those who had passed, opinions were nearly evenly split.
- The respondents’ perspectives about pros/cons of AI utilization depended on their evaluation of human capability.
Taking that data into consideration, we picked 4 categories; two of them ("Medical care" and "Disaster prevention") are of particular interest, and the others ("Love" and "After death") are controversial and polarized among respondents. Then we visualized their opinions in the form of map.
(Note: Hereafter, we will use the term “Digital cloning of the deceased” or “Digital cloning” instead of “After death” on the map, because “Digital cloning” better describes the situation.)
The perception of AI differs greatly according to the categories.
How does the willingness to use AI differ among four categories? Let's take a look at the below pie charts.
While over 90% of respondents were eager to use AI for “Medical care” and “Disaster prevention”, this figure dropped to 60% for “Love”, and less than 50% for “Digital cloning”. It can be seen that the sentiment toward AI differs greatly according to the categories.
Opinions on AI
Grouping respondents’ opinions, we found that they can be categorized into 4 groups (zones); Trust, Expectation, Rejection and Anxiety.
(1) Zone: Trust (upper left)
Opinions in this zone may indicate “AI is something reliable to overcome confronting challenges”.
- Medical care: “Easy access to medical advice”
- Disaster prevention: “Judgment based on the situation”
- Love: “We do not have fair eye on love”
- Digital cloning: “It will support me mentally”
(2) Zone: Expectation (upper right)
Opinions in this zone may indicate “AI will lead society in a positive direction”.
- Medical care: “No more shortage of doctors”
- Disaster prevention: “The level of accuracy will be higher”
- Digital cloning: “I can talk to deceased people from the distant past”
(3) Zone: Rejection (lower left)
Opinions in this zone may indicate “AI cannot replace humans”.
- Medical care: “I am more comfortable with human doctors”
- Love: “I do not think love can be measured in data”
- Digital cloning: “It will ruin my/our memory”
(4) Zone: Anxiety (lower right)
Opinions in this zone may indicate “AI would have a bad influence on society”.
- Disaster prevention: “Less attention to disaster prevention”
- Love: “It seems that AI controls my feelings”
- Digital cloning: “Potential harm in politics”
The relationships among these four zones are displayed on the above chart.
The vertical axis denotes sentiments toward AI (Positive or Negative), while the horizontal axis denotes if the opinions are headed for personal confronting issues (Intuitive) or social issues (Objective).
Relationships between Positive and Negative opinions
A remarkable point is that some of the positive and negative opinions refer to close to, or the same matter. Please look at the AI Map. The summary of some respondents’ voices (hereafter, “opinion”) are in the circle, diamond and hexagon, and it is allocated next to one representative respondent’s voice (hereafter, “voice”). Some positive and negative opinions are joined by dashed arrows. On the below piece of map, for example, you can see the opposite opinions on the same matter.
Expectation and Anxiety, Trust and Rejection - Each of these opposing sentiments show both positive and negative effects of AI on society.
How to view this map
Now, let’s look at the map again. You are free to browse the map as you wish, but allow us to introduce one example of how to view this AI Map.
Step 1: Choose a category that interests you most from "Love” (Pink), "Medical care” (Teal), "Disaster prevention” (Green) and "Digital cloning” (Purple).
e.g.) This time, suppose you choose “Love”.
Step 2: Find opinions that interest you most while following the same color.
e.g.) From the category “Love”, you choose the opinion “I do not think love can be measured in data”.
Step 3: Trace the line from the opinion you choose and look at the other opinions and voices.
e.g.) Tracing this line, you may find opinions with negative perspectives like “Advice from others are not needed” and “Advice from AI is just a reference, not more than that”. It can be guessed from these opinions that the ultimate decision should not be done by AI but by humans.
In contrast, you may find that the positive opinions are linked to the negative ones. For example, “I do not think love can be measured in data (, so I do not want to use AI)” is connected to “We do not have fair eye on love (, so I want to use AI)”. Both of those are true, and may depend on individual’s values. Which opinions would be the closest to yours?
As you continue, you see other opinions, such as “I do not have anyone to talk to about my love”, “I do not have a lot of experiences”. This reflects the diversity of users.
As such, selecting one opinion as a starting point and tracing relevant opinions, you may be able to get a chance to discover new ways of thinking, or to reflect on and refine your ideas. We hope that you can use this AI Map in this way.
It is YOU who will complete this AI Map.
The opinions on this map are just a small representation of society’s opinions. Together with your colleagues, friends and family, look at the AI Map and discover who agrees or disagrees with each opinion. The discussion may become a good chance to find different perspectives that you were unaware of. Then talk about how you developed your own perspectives. This will be your own, original AI Map.
The attitude to be open to others’ perspectives will lead to the development of an AI-ready society, because the AI is developed by people, for people. Communication will enrich our relationships with AI in the AI-driven society.
Who are the respondents of this survey?
The survey was conducted at Miraikan’s permanent exhibition floor and on the website.
You can see and answer the questionnaire from the below link.
“Hello! AI World - What do you want to do with artificial intelligence? -”
Survey period: 17:00 on February 3 to 17:00 on February 24, 2020
Survey method: Opinion Bank (5F on Miraikan’s permanent exhibition floor and Website)
Surveyor: Miraikan - the National Museum of Emerging Science and Innovation
Cooperation: The Japanese Society for Artificial Intelligence
Valid data: 887
- Participants by gender (SA)
- Participants by age (SA)
- Level of comprehension of AI (SA)
Credits of “AI Map: From Everyone, for Everyone”
Production: Miraikan - the National Museum of Emerging Science and Innovation/ Miraikan Focus Project Team (Fumiya Urushibata, Yoshiyasu Watanabe, Ryu Miyata, Ayuko Sakurai, Atsushi Ozawa and Mizuki Kawano)
Cooperation: The Japanese Society for Artificial Intelligence
Design: Takashi Tokuma (bowlgraphics Inc.)
Illustrator of the questionnaire: Mary Tomita
"AI Map: From Everyone, for Everyone" website (Japanese)
Opinion Bank website (Japanese)
Opinion Bank website (English)
Project “Miraikan Focus” website (Japanese)