ChatGPT "Another Dimension" AI ShockApril 4 at 11:15

ChatGPT is an AI that learns a huge amount of information on the Internet and can talk like a human.

It is rapidly spreading around the world and is beginning to have a major impact on our lives.

"It's a revolution on par with the invention of the Internet and mobile phones" (Bill Gates) "It's terrifyingly amazing, dangerously powerful AI has become a reality" (Elon Musk)


"Another dimension" of AI in which both positive and negative aspects are greatly undermined.

"How will we humans deal with AI?"

The inevitable proposition is now confronted.

Elementary school students write

"As I read on, I was drawn into the story and couldn't let it go until the end," "I sympathized with his attitude of trying to protect the protagonist, and I couldn't help but weep at his fate,"


a passage from a reading review of the Harry Potter series submitted by a fifth-grader last month.

The handwriting is appropriate for his age, but the expressions used are so skillful that it is hard to believe that they were written by elementary school students, such as "I was impressed" and "I cried."

It was written using "ChatGPT".

The homeroom teacher who received the submission of the essay recalled:

The homeroom teacher
at school: "I was surprised because it was well structured and the way it was written was the kind of expression used by adults. However, just copying the writing style is a learning experience, and I appreciate their willingness to incorporate new things."

The use of ChatGPT is rapidly spreading Japan and is beginning to change our lives.

Revolutionize your business and make significant operational improvements

In February, Panasonic Connect, an IT company that develops systems for enterprises, introduced a unique AI system using ChatGPT to all 2,1 employees in Japan.

He says that he has already seen dramatic effects on operational efficiency.

Employee
: "I use it to create scripts for presentations and to create meeting minutes, and they also provide colloquial English translations, which makes it easier," and "They answer any questions anyway."

IT employees analyzed more than 1000,1 surveys they conducted asking employees about their satisfaction with the work environment.

It takes a lot of time to decipher the description written in the free description column about what each person is tasked with. Instruct them to write a program that automatically evaluates their content.

If you type, "Please create code that analyzes the sentiment of a specific item of data imported in Python and repeats the process until all records of the imported file are processed."

In a few seconds, you have a draft of the program's code.

It is now possible to create a program in 1 minutes, which used to take one hour to write.

An employee
of the IT department: "It comes out quite accurately and with a very clean syntax, so you can use it as it is. I think it will speed up dramatically."

In addition, employees of the Legal Department used it to prepare materials and come up with ideas for use in in-house training.

When asked to create quiz-style teaching materials to help students learn about the Subcontract Act, the AI instantly presented four options as questions and corresponding answers, and also showed explanations.

"I feel that ChatGPT gives me another idea" At the company, AI is only an assistant, and it has decided on "5 precautions"

when using it so that it should be used after understanding its limitations.

"It's not always accurate, the final decision must be made by people."
"We can't answer recent questions,"
"We can only answer from public information. We haven't learned any internal information."
▽ "English gives more accurate answers" ▽ "I don't know about the future"


The person in charge who decided to introduce it expects that the productivity of work will be further improved by using the spare power and time generated by the efficiency of AI.

Akihiko Kono, Executive Officer and CIO
: "I think that each employee has been able to use it in a variety of ways than we expected, and I think that the efficiency improvement will allow us to shift to spending time on judgment and brushing up."

Jobs are transformed, and employment concerns are also

In the United States, where ChatGPT is becoming more widespread, the sheer power of ChatGPT is causing ripples.

Oded Batat, owner of a travel company in Connecticut for nearly 50 years.

I asked ChatGPT for a travel plan proposed to the customer.

I asked for a trip to Paris and a budget of $3,5000 for three people. Then, a plan of 4 nights and 5 days was immediately proposed, which would allow you to efficiently visit famous sightseeing spots such as the Louvre.

Travel agency owner Oded Batat
said: "It would be amazing if it could be made more sophisticated. However, if you keep up with the times, it may be difficult to keep working. Maybe it will cause travel companies to decline or take away jobs."

As for the impact on employment and jobs, last month, a research group from the University of Pennsylvania and others in the United States published the results of a study on how AI such as ChatGPT affects people's jobs.

It was said that "the work of nearly 80% of people in the United States will be affected in some way."

Science fiction magazine floods with applications

In some places, the impact is becoming apparent.

Neil Clark is the editor-in-chief of an American science fiction magazine. The magazine has produced famous literary awards, but the number of posts that appear to have used ChatGPT has increased dramatically this year.

Until now, there have been only a few works a month that are thought to have been written by AI, but in February we received 2 submissions.

Neil Clark, Editor-in-Chief
: "It all started when YouTube and TikTok talked about how to make money using ChatGPT. My magazine became one of the targets."

As a result, Clark was forced to suspend the call for submissions.

Neil Clark, Editor-in-Chief
: "It's depressing, if you don't accept new work, you can't innovate or change. Some of my peers say, "I don't open the call for works anymore." Every time the door closes, the entire industry is hit."

Significant impact on education

Education has had a major impact.

Anthony Orman, a philosophy professor at Northern Michigan University in the United States, is wondering how to respond to the ChatGPT that students use.

When we talked to the students, many of them were positive about using it.

(Female student)
"I use ChatGPT to write reports, and they help me. They are very smart and knowledgeable, and they will be able to make much better things than if they were made alone."
(male student)
"I will continue to use it, what is the difference from using Google? Even without ChatGPT, some kids will cheat."

On the other hand, teachers are increasingly calling for a ban.

As for my fellow professors.

"I think it should be banned, I teach history, but I want you to read and think about the material yourself. I think it makes sense to do it on your own."

Should it be banned or not?

Professor Ohman said he had not yet made a decision.

Professor
Anthony Orman: "I'm very conflicted, and I want to use it. It's going to be the norm from now on. On the other hand, I am also worried that what I have done so far will be ruined."

Measures in the field of education Japan

教育現場での利用は日本でも広がっていると見られるが、詳しい実態は分かっていない。

こうした中、国内の大学でも対応や見解や注意喚起を始めるところも出てきた。

東京大学は、学内向けのホームページに副学長名で見解を示し、「学位やレポートについては生成系AIのみを用いて作成することはできない」などとした。

また、上智大学は「レポート、小論文、学位論文等において、使用を認めない。使用が確認された場合は、厳格な対応を行う。教員の許可があればその指示の範囲内で使うことは可」などとする方針を示している。

東北大学では「未発表の論文や秘密にすべき情報を入力してしまうと、意図せず漏えいしてしまう可能性がある」などとする、注意喚起を行った。

このほか、小中高校生の「読書感想文」の今年度のコンクールについても、主催する全国学校図書館協議会が、応募要項に「盗作や不適切な引用等があった場合、審査対象外になることがある」と追記したほか、文部科学省も、今後、学校現場での取り扱いを示す資料を作成する方針だ。

大きな脅威も

驚くほど進化が速いAI。その脅威に世界中で懸念の声が高まっている。

アメリカでは、非営利団体が、高度なAIの出現で人類が文明を制御できなくなるおそれがあるなどとして少なくとも半年間は開発を中断するよう求める書簡を公表。

実業家のイーロン・マスク氏やアップルの創業者の1人スティーブ・ウォズニアック氏など著名人ら、11日時点で2万人を超える署名が集まっている。

データ利用の問題は

懸念されていることの一つは、AIが学習するインターネット上のデータ利用の問題だ。

データには、個人情報やビジネスの機密情報なども含まれる可能性がある。

イタリアの当局は、個人データを法的根拠がないまま収集し、利用者の会話の内容や支払いに関する情報についてデータの侵害があったとして、個人情報の保護に関する法律に違反している疑いがあると一時的に使用を禁止すると発表した。

これに対して、開発企業は、AIが学習するデータには個人情報も含まれているものの可能なかぎり削除している、AIが作成する回答は、暴力などを含む内容は生成されないようにしている、としている。

また利用者は18歳以上、または保護者の許可を得た13歳以上に限り、今後、年齢を確認する仕組みの導入を検討しているとしている。

フェイク生成 サイバー犯罪への悪用は

もう一つは、誤った情報やフェイクニュースの拡散、国家のプロパガンダなどに利用されるなどのリスクだ。

特定の国や政治家などに好意的、または悪意をむけるような記事を量産すれば世論をコントロールできてしまうおそれもある。

サイバー犯罪に悪用されるリスクも指摘される。

アメリカの調査会社が公表した、2023年の世界10大リスクではロシア、中国に次いで3位に「大混乱生成兵器」というタイトルで、ChatGPTなどの生成系AIの進歩や普及が、政治・経済的な混乱に広く影響するおそれがある、などとしてランクインした。

こうした、AIがもたらず脅威に関して、EUでは世界初のAI規則案の制定を目指している。

東京大学の松尾豊教授は、ルール作りの必要性を指摘する。

Professor Yutaka Matsuo of the University of Tokyo:
"Since it has the potential to have a significant impact on human decision-making, we should create rules such as 'using it this way is good or bad' as soon as possible.

Exclusive interview with the developer

What do the companies that developed ChatGPT think?

CEO Sam Altman said in an exclusive interview with NHK:

"What developers need to do is think about how to deal with the risks and make sure people can reap the big benefits. This new technology will enable amazing new things and enrich our lives in ways that are unimaginable right now. AI has been portrayed as robots that enslave humanity in science fiction for decades, and it's easy to think of it that way. But we make it so that humans can set rules and stop them. Regulatory discussions are very important. The government has an obligation to discuss it, and it is willing to participate in it. Technology has great advantages and serious disadvantages, and the Industrial Revolution said the same thing. AI enhances our creativity, not replaces [humans]."

A future with AI

The day will surely come when "a society living with AI" will become commonplace.

At an elementary school in Takahata Town, Yamagata Prefecture, classes have begun to learn how to interact with AI by incorporating ChatGPT.

On this day, it was a morality class, and the theme was "To rely on AI or not to rely on it." The teaching material used is a story about a boy named "Kazuya-kun" who is nicknamed "Side Dish" by a good friend, and although he doesn't like it, he can't confide in his worries because he doesn't want to be hated.

The people around him gradually began to call Kazuya-kun a side dish, and Kazuya-kun became more and more burdened with his steps to school.

The teacher consulted with ChatGPT about Kazuya's problems. And then...

ChatGPT
: "The first thing you need to do is tell your friends you don't like the nickname, and if that's not right, you can write them an email or letter to let them know how you feel."

The teacher asked the students what they thought if Kazuya would have relied on ChatGPT for this story. Opinions diverged.

"I rely on AI, because it will teach me a lot and give me solutions," and "I will go to my mom or dad first, and if it doesn't work, I will consult with AI."



"I don't rely on them, because I thought it would be better to solve them ourselves so that I could convey my feelings," "I don't rely on them. Because if you rely on AI, you may rely on AI for the rest of your life."

In class, the teacher asked the children many questions.

"Who's the last to judge?"

Doraemon Country

Some of the ChatGPT development teams were actually born in Japan.

Technical executive Shayne Gu. Regarding the way Japan people interact with ChatGPT, he described it as interesting, unique and deep, citing Doraemon as an example.

Shayne Gu:
Unlike other countries, Japan are not afraid of AI or see it as a tool, and I think that is largely due to the influence of Doraemon. From an early age, he somehow understood what it was like to live in a world with AI. I think this is unique in the world. I think Japan has the greatest potential for growth in the field of generative AI."

Skynet or Doraemon

The more I report, the more I report that "humanity will be dominated by AI" and "AI can solve many of the problems that the earth is facing,"

the more I am exposed to complex thoughts and feelings that are a mixture of anticipation and anxiety, and the more I am at the mercy of them. AI is evolving so fast that even experts say it's becoming harder to get an accurate picture of what's happening around the world.

Innovative technology has always been welcomed and embarrassed.

It was the same with the Internet and smartphones.

Whether the Internet and smartphones have enriched us or made us happy, there are no results.

And this ChatGPT. What is being tested this time is humans again. I think it's our way of life.

Will it be "Skynet" that destroys humanity or "Doraemon", your best friend?

It is up to us to choose that future.

Kenichi
Nikijima, Director
of the 2nd Production Center (Science), joined the station in 2014 and is in charge of many science programs related to AI and the
new coronavirus.

Production Center 2 (Science) DirectorMidori
AmanoJoined
the station in 2021 In charge of many science programs such as
Science ZERO Hobbies are stargazing and collecting minerals and fossils

Reporter
Naoaki Shimada of the Science and Culture DepartmentJoined
the Fukuoka Bureau
in 2010 and is now a member. In addition to digital fields such as AI and metaverse, he is in charge of a wide range of scientific fields such as biology

Los Angeles Bureau Reporter
Nana
YamadaJoined
Nagasaki Bureau in 2009After working in the Economic Department and the International Department, he now belongs