Ryan General
Ryan General3141d ago

Japanese AI Robot Takes University Entrance Exam, Exposes Alarming Flaw in Human Education

Japanese AI Robot Takes University Entrance Exam, Exposes Alarming Flaw in Human EducationJapanese AI Robot Takes University Entrance Exam, Exposes Alarming Flaw in Human Education
A robot with an advanced artificial intelligence bested 80% of human students in the recent Tokyo University entrance exam.
Interestingly, not only did it reveal what can still be improved about AI technology, it also exposed a flaw in human education that needs to be addressed, an AI expert pointed out.
The machine, dubbed as Todai Robot, is part of a project by Japan’s National Institute of Informatics that aims to understand the types of intellectual abilities machines will be able to replace in the labor market. In 2011, the program initiated the development of Todai, an AI that was supposed to be “smart” enough to enter Japan’s most prestigious university.
While Todai failed to achieve passing scores in Tokyo University’s 2016 entrance exam to gain admission, it did perform in the top 20% of exam takers, according to Business Insider. The score is also enough to enter more than 63% of the universities in Japan.
At the 2017 TED Conference, project director Noriko Arai explained that instead of being impressed, she found it alarming that the robot answered questions correctly without actually understanding a thing. Arai pointed out that while the result was able to show that machines indeed have the potential to replace a huge part of the global workforce, it also highlighted a flaw in human education.
Arai observed that just like her Todai robot, which fails to absorb meaning, human children also ingest facts without comprehension. Todai, which has 15 billion sentences in its memory, still failed to answer a multiple choice question whose answer would have been obvious even to young children. However, in Arai’s follow-up tests, she found that easy questions can sometimes also stump humans.
As humans, we can read and understand, which Arai noted computers cannot do yet. That is why we are more adept to problem-solving, pattern recognition, and doing creative endeavors.
However, the research revealed that AI will eventually surpass human memory and cognition in the near future, and humans not being able to keep up with machines should be a cause for concern.
To prepare for a future wherein AI exceeds human capacity, Arai figured that a better education system is needed. She suggested an education in which kids are taught not just to collect and memorize facts, but also to analyze and think critically about them.
“So we have to think about a new type of education,” Arai said. “How we humans will coexist with AI is something we have to think about carefully, based on solid evidence. At the same time, we have to think in a hurry because time is running out.”

Discussion

Ari C.
Ari C.2h ago

If this happened on campus, Stanford should issue a clear public update and specific safety actions.

212 Face
Mina Z.
Mina Z.1h ago

Agree. People need facts and process, not silence. The school should confirm what is being investigated.

88 Face
Ken L.
Ken L.48m ago

Also important to separate verified details from rumors so this does not spiral online.

61 Face
Linh P.
Linh P.1h ago

The death threat part is extremely serious. Hoping law enforcement and campus security are already involved.

144 Face
Jae T.
Jae T.35m ago

This is where official reporting and support channels need to be visible and easy to access.

42 Face
Sophie W.
Sophie W.56m ago

Can NextShark keep a timeline thread here as updates come in? That would help keep context in one place.

97 Face
Your leading
Asian American
news source
NextShark.com
© 2024 NextShark, Inc. All rights reserved.