Can Artificial Intelligence really be made? Part 2 of 3

, Chatani Masayuki, AI

[Artificial Intelligence also challenged to Quiz Champion]

Artificial Intelligence had challenged Quiz Champion. AI fought against on the famous TV quiz program “Jeopardy” in United States. The AI fight was “Watson” created by IBM. This quiz show has a style to answer as quickly as possible and type of question is known like “The capital of the country is Tokyo”. Every round there are about six topics and the difficulty of the quizzes increase as the reward amount becomes higher.

Since “Jeopardy” has not repeated the past quizzes therefore it is reported that they did not take an approach to memorize the past quizzes but also take an approach to learn many articles. This really exciting story of development is introduced in detail in a book “Final Jeopardy: The Story of Watson, the Computer That Will Transfrom Our World (Japanese title : IBM 奇跡の“ワトソン”プロジェクト:人工知能はクイズ王の夢をみる).

[Artificial Intelligence that even creates recipe of cuisine]

After the battle with quiz champion, Watson is considered to introduce medical, healthcare and other domains by taking advantage of its mechanism. By learning a lot of treatment data, Watson intends to improve the diagnostic accuracy by estimating possible disease, indicating medical check to additionally take. Watson is applicable to the domain where correlation data like symptoms and diseases exists therefore it is expected to apply to domains other than medical and healthcare.

Interesting application with Watson, it is a project called “Chef Watson” and it is an initiative to create new cuisine recipe with the technology of Artificial Intelligence. Watson learns and studies various recipe from cooking books and magazines and how to cook and tries to propose never-seen recipe from the infinite numbers of material combinations. It is paid attention as new application domain for Artificial Intelligence since a recipe book “Cognitive Cooking with Chef Watson” was published in 2014.

[Toyota enters the domain of Artificial Intelligence]

Early 2016, Toyota delivered a big news in Artificial Intelligence domain. They invited a researcher who ran autonomous driving car at Google and established a subsidiary for Artificial Intelligence research “Toyota Research Institute”. Dr. James Kuffner who is invited this Toyota Research Institute is reported to manage “Cloud Rotobitcs” which provides recognition technologies required for robot from the cloud.

Toyota also accelerates the investment in mobility area of Artificial Intelligence and announced collaboration with Preferred Infrastructure, a research venture of Japan.

Toyota advocates one-person personal mobility and tackles the problem of mobility in the long-life age and assumes to introduce autonomous driving technology with Artificial Intelligence.

[Is mankind ruled by Artificial Intelligence?]

As we see Artificial Intelligence being popularized and generalized into society, we hear some voices of concerns against its penetrations into our life. For example, Dr. Steven Hawking, a well-known astronomical physicist, Mr. Eulon Mask, Chief Executive Officer of Space X for space transportation business as well as Tesla, an electronic automobile and Steve Wozniak, known as Woz, one of founder of Apple. They all show their concerns to the potential war caused by autonomous weapon with Artificial Intelligence equipped.

At the event “Zeitgeist 2015”, Dr. Steven Hawking, astronomical physicist, had warned that “Within 100 years, intelligence of robot may terminate the human civilization” and after that various opinions and thoughts are provided and exchanged from many experts.

It means an autonomous type is some mechanism where Artificial Intelligence understands situation, makes its own decision, acts by its own risk. Android (Humanoid) T-800 performed by American actor and former governor of State of California, Arnold Schwarzenegger and the latest version T-1000 made with liquid metal are the example of embedded Artificial Intelligence in the android.

Previously mentioned DeepMind, a company Google acquired, it is reported that both companies agreed not to use DeepMind technology for military purposes as a condition to be acquired and this shows the fact that many Artificial Intelligence experts have concerns Artificial Intelligence may be equipped on weapon and become uncontrollable.

[Artificial Intelligence everyone already use]

In fact, many of you have already experienced with so-called Artificial Intelligence. For example, in the video games, there is a technology domain called Game AI for a long time. When play sport game like Soccer / Football, you are controlling the move of player through your input from the game controller.

Game AI generates the movements of all of players in the field no matter it is your team or opponent team other than you control based on the computation in the game applications. Character other than player is called NPC or Non Player Character and NPC determines its actions in the game autonomously and behaves with Artificial Intelligence.

For example of Japan team (or any national team) in the Soccer game, respective field player is designed to represent real physicals and play styles and formation control is beautifully executed to be shown as a team of real Japan national team. This game AI domain had made a tremendous progress and it offers the great experiences like being one of players of national team or famous club team.

(Continues to Part 3 of 3)