Waseda

About the Use of Generative Artificial Intelligence (ChatGPT, etc.)

• 发布单位Publishing unit Waseda University / Senior Executive Vice President (Academic Affairs/Provost)
• 官方来源Official source https://www.waseda.jp/top/en/news/77786
• 页面日期Page date 21 Apr 2023

---

About the Use of Generative Artificial Intelligence (ChatGPT, etc.)

To all students, faculty and staff

April 21, 2023

Waseda University

Senior Executive Vice President (Academic Affairs/Provost)

Koichi Suga

About the Use of Generative Artificial

Intelligence (ChatGPT, etc.)

I ntellect ual resilience and flexib le sensitivity

Waseda University has been asking students to train themselves to develop “intellectual resilience” so that they can think thoroughly about problems with no correct answer and come up with their own answers, and to foster “flexible sensitivity” that allows them to work together with a variety of people of different races, nationalities, religions, ethnic groups, languages, genders, and sexual orientations in order to solve vari ous problems with respect and sympathy for one another. In today’s world, where we face major issues such as pandemics, wars, and progressing technological innovation, which can be said to be turning points in society, the importance of “intellectual resilience” and “flexible sensitivity” is increasing.

For example, th e development of new technology will change the way we live, work, and the state of society.

It’s an unavoidable reality that scientific progress and technological innovation will always have positive and negative sides, advantages and disadvantages.

It de pends on how humans use technology, and on national policies and institutions that promote or regulate these technological innovations.

For those of you who aim to strengthen your “intellectual resilience”, nurture your “flexible sensitivity”, and contribut e to society, realizing the correct use of new technology is an extremely important issue.

It will also be an important task to conduct research on measures to mitigate the social impact and disseminate the results, with a “flexible sensitivity” that sympa thizes with those who are negatively affected by new technology.

Fundamental Attitudes towards Generative AI

Generative artificial intelligence (hereinafter referred to as generative AI) such as ChatGPT

(OpenAI), Bard (Google), and Bing (Microsoft), wh ich have become a hot topic in recent years, are advancing their performance at an accelerated pace.

It has been pointed out that they have the potential to bring about major changes in our society. For this reason, interactive automated response services based on generative AI are not only tested around the world but also discussed at educational institutions as to how they should be utilized and regulated.

Those who advocate “intellectual resilience” need to be familiar with the correct usage of generat ive AI.

To that end, we must have sufficient knowledge of generative AI and cultivate an attitude of keeping pace with technological progress and continuing to innovate our knowledge.

Because then you can masterfully use the technology instead of being the one used by it.

In addition, it is imperative to have an accurate understanding of the impact that generative

AI will have on society and people, and the issues that arise around the formation process, when we consider the issues of whether to stop the development or regulate the use of generative AI, all of which are in the midst of debate.

This is because in the development of new technologies and the production of products and services based on them, a great impact on various people is observed not on ly after development and production but also before.

Characteristics of Generative AI

Generative AI learns by itself based on the vast amount of data on the Internet, and can instantly generate response sentences in a natural language dialogue format f or various questions and requests in a wide range of fields.

Also, even if you ask the same question, the answer (sentence) that is returned will change each time. The following tasks that were previously performed by humans can now be handled by generative AI.

(1) Generation of new sentences based on given string s and keywords (creation of reports, impressions, etc.)

(2) Generating answers for tests (fill

– in

– the

– blank questions, multiple

– choice questions, open

– ended questions, etc.)

(3) Brainstorming, creation of ideas

(4)

Summarization, translation and proofre ading of existing texts and web pages

(5)

Data analysis

(6)

Easy programming

(7)

Lyrics, composition

(8) Searching for documents and information

If you can use generative AI with these functions well, you will be able to save a lot of your work time, and many people will get the benefit of being able to concentrate on other important tasks.

On the other hand, the following flaws and difficulties have also been pointed out.

There are many things that students need to be careful about.

First, genera tive AI does not always output accurate information.

Currently, f or example, there are quite a few cases where answers are generated using non

– existent universities or companies, incorrect numerical data

, or wrong names.

It has also been observed often to fabricat e bibliographies of titles and author names that do not exist.

Second, it has been pointed out that reports generated by generative AI are (currently)

logically correct as text, but the content is shallow and there is a tendency for a repetition of plausible and obvious things.

If you use the results of the generative AI without examining the content, the report will be lacking in originality.

Third, if you submit the thesis created by the generative AI without carefully examining it, the thesi s will be evaluated as it is, even if there are incompleteness or fatal errors in the content.^{核实责任Verification duty} You may also be subject to disciplinary action if plagiarism, inappropriate citation, or fabrication is confirmed in the submitted thesis, even if you did not in tend to do so.

Of course, if you submit a thesis created by the generative AI as it is, it will be punished as a fraudulent act similar to cheating.^{违规定性Integrity framing}

Fourth, careless use of generative AI may lead to the leakage of confidential information within the univ ersity or company, and questions that include your name, etc. may lead to the leakage of personal information.^{数据与隐私Data & privacy}

It has also been pointed out that the disorderly use of generative AI can lead to discrimination, human rights violations, and criminal behavior.

Freedom of action must be guaranteed for technological progress, but no one has the freedom to infringe on human rights.

“Intellectual resilience” must also pay attention to the tension between technology and society.

Usage and Limitations of Generative

AI

Surely you might be able to complete a report quickly with generative AI, but you’re giving up the opportunity to improve your abilities.^{学习理由Learning rationale摘录Excerpt} Reports are created by gathering a large amount of information from documents and materials, classifying, analyzi ng, and summarizing them to create the most appropriate answer to the task.

Writing a report is the most important exercise in developing

“intellectual resilience”. As a result of neglecting this task and prioritizing short

– term profits, in serious situations such as entrance exams and work in the real world, you will notice the fact that you have learned nothing.

On the other hand, for those working as business creators who envision unprecedented products and services, the use of generative AI will lead to the realization of effective work.

In the process of collecting, classif ying, summarizing, and analyzing vast amounts of information about similar past cases to generate new ideas, the partial use of generative AI could save time, and y ou could spend more of your time generating important new ideas.

This effective use assumes that you know firsthand how much work is needed and what quality is required at each stage in the process of making a report. That way, you can easily determine the merits of substituting that part with generative AI, as well as discrepancies and mistakes in the responses shown by AI, and you can use it in a suitable manner. It is important to train your

“intellectual resilience” daily so that you can use generative AI appropriately according to the time and situation.

In addition, there are reservations about generative AI in terms of “flexible sensitivity”. To be able to work together to solve social problems while respecting the diverse attributes and backgrounds of various people, it is necessary to build mutual trust.

It requires both mutual understan ding and sympathy.

A certain kind of expression is also required in exchange of messages, but it seems difficult for generative AI to appropriately provide expressions that show fine

– grained correspondence and persuasive power.

Future Action

You will c ontinue to utilize a variety of products and services developed based on new technologies.

Before using them, or while using them, you should repeatedly think about whether they are necessary for you, whether they are beneficial to you, and what kind of im pact they have on society.

What is beneficial to you now may only have a negative effect on your future self.

Some people get more benefits than disadvantages, others do not. Think objectively and with an altruistic spirit.

Waseda University’s students, fa culty, and staff, as well as all concerned parties, should not forget this way of thinking.

“Intellectual resilience” applies to all matters in the world.

To fully understand the multi

– faceted impact of new technologies, how to curb the negative aspects, and to further expand the positive aspects, we (not only individuals, but also governments, industries, communities, etc.) must think carefully about what we should do.

AI and the Future of Humanity

• 发布单位Publishing unit Waseda Weekly
• 官方来源Official source https://www.waseda.jp/inst/weekly/features/specialissue-ai/?lng=en
• 页面日期Page date 2023-06-21

---

Since the end of 2022, generative artificial intelligence (hereinafter referred to as generative AI), including "ChatGPT", has been attracting a lot of attention. ChatGPT is an interactive AI that when you type what you want to ask into the input field, it returns an answer as if you were talking to a person. Due to its high level of convenience, it is attracting attention as a new tool that can replace traditional search engines.

How will this AI change our lives and our future? Also, how should we deal with AI? This time, Waseda University students actually experienced ChatGPT and spoke to Professor Tetsuya Ogata and Professor Daisuke Kawahara of Waseda University's AI Robot Research Institute.

What kind of tool is "ChatGPT"? A Waseda University student actually tried using it.

ChatGPT (Chat Generative Pre-trained Transformer) is an interactive generative AI released by OpenAI in the United States in November 2022. Currently, GPT-4 has been released to the public after many version upgrades, and its performance is increasing.

In response to growing interest in generative AI such as ChatGPT, Waseda University has officially announced its position on the use of generative AI

.

What kind of generative AI is it that is attracting so much attention? First, we asked Shino Watanabe, a Waseda Weekly reporter and second-year School of Education, to try out ChatGPT.

Today is my first time using ChatGPT, but some of my friends and seniors around me seem to be using it. In particular, when I heard that it was being used to create entry sheets for job hunting, I was surprised and wondered, "Is it really being used that practical?"

I'm also interested in ChatGPT, so I'm looking forward to trying it out! I'll immediately ask ChatGPT the three questions I've been thinking about.

What is the significance of university?

As soon as I typed and submitted my question, ChatGPT started generating text! The text comes out as if a person was typing. It's kind of strange.

Universities have various meanings. The main meanings are listed below.

1. Acquisition of knowledgeAcquisition of knowledge University is a place to learn specialized knowledge, and you can deepen your knowledge in the field that interests you. This will help you in your future career and personal growth.

2. Research and discoveryResearch and discovery Universities are also places for the discovery and development of new knowledge and technology. Students and Faculty conduct research and contribute to the development of society as a whole by bringing about new discoveries and technological innovations.

(Original text)

I'm sure. I will omit it here, but there is actually a detailed explanation in the following sections.

3.Improvement of logical thinking and problem-solving skills

4.Building relationships

5. Gaining a global perspective

6.Social contribution

(Excerpt from original text)

The answers are convincing, and it's amazing how they can be summarized in such a logical and easy-to-understand manner.

Let me ask you the following question.

Please tell me the latest information about Waseda Weekly.

I'm sorry, but my knowledge is up to date as of September 2021. Therefore, we are unable to provide the latest information on Waseda Weekly.

(The following is omitted. Original text)

that? It seems like you can't answer that. After looking into it, it seems that ChatGPT cannot currently support information after September 2021

.

So, next is this question.

What is the most important thing in communication?

The most important thing in communication is "respect for the other person." (Omitted) Respect-based communication includes the following elements:

1. Active listening

2.Open-minded

3. Respect the other person's feelings and position

4. Clear communication

5.Body language

(Excerpt from original text)

Hmm, that seems to be the case. Although I have a weakness for up-to-date information, I'm surprised that I can get so many answers so quickly.

How will generative AI like ChatGPT change the world in the future? I've been curious about a lot of things, so this time I'd like to talk to Professor Tetsuya Ogata and Professor Daisuke Kawahara of Waseda University's AI Robot Research Institute, who are experts in AI!

What will the future hold for AI and humans? Interview with two Professor from the AI Robotics Institute

What are the impressions of the two Professor who saw the answers on ChatGPT?

This time, I asked ChatGPT three questions, including the significance of university. What do you think after looking at the answers?

Your answer is very logical and clear. It's a little too safe though (lol). But it

's actually amazing to be able to get a safe answer without fail

. In the past, there have been AIs that have caused problems due to their extreme or biased answers. I think ChatGPT is adjusted so that it will give you acceptable answers.

However, perhaps because ChatGPT only references information up to September 2021, it did not respond to the latest information on Waseda Weekly.

I wish I could answer that (lol). This can be said to be a weakness of ChatGPT at the moment, but it will be addressed eventually.

In fact, Microsoft's search engine "Bing" uses ChatGPT technology, and it seems that it can also provide the latest information.

ChatGPT is also interesting because even if you ask the same question again, you will get a different answer. Maybe there is a parameter (*3) that makes the answers vary each time.

(*3) Values that affect the mechanisms of machines, systems, software, etc. from the outside.

In that sense, they are similar to humans. Humans also give different answers to the same question at different times.

That's right, it's the context. ChatGPT takes into account the information exchanged before the question and returns an answer that is appropriate at that time. When you take into account the context of the conversation, you suddenly start to feel more human.

Generative AI will become essential to future society

Teachers, what kind of impact do you think generative AI like ChatGPT will have on society?

This is already happening, and I think it will definitely have a huge impact on society

. If you use it to support your work or studies, you can be more efficient, and you can also brainstorm ideas. In private, you can use it as someone to chat with, and there are many ways to use it. Many people are probably thinking about how to make use of generative AI right now.

On the other hand, I think you should be careful because you can do so many different things. For example, it would be better to think that information on the internet will basically be referred to by generation AI

. I think some people may feel uncomfortable about that.

Is there a possibility of abuse?

Yes, there is. Cybercriminals may try to have other generative AIs create programs for computer viruses, which I believe ChatGPT is not designed to do.

Of course, I think companies developing generative AI are trying to avoid such risks. However, the field of AI is still in its infancy, so some people may be looking for a loophole, and it may become a game of cat and mouse.

I think the scary thing about this is that an atmosphere that generates AI should be regulated becomes excessively strong in the world

. Things have improved now, but in the past there were excessive regulations that made it difficult for drone research and development to proceed as it was considered dangerous if a drone crashed.

To prevent the same thing from happening to generative AI, I think it is necessary to proactively listen to the opinions of legal experts and make appropriate regulations.

The benefits of generative AI are immeasurable, so it would be a big problem if it became unusable.

Will generative AI like ChatGPT become essential to society in the future?

I think it will become indispensable, just as we can no longer part with our smartphones

. Generative AI will improve efficiency in all fields, and simple tasks will no longer be the work of humans. As a result, I think we will become a society in which people are more creative and focus on the things that really matter

.

Use AI effectively with "critical thinking"

Generative AI has its advantages and disadvantages, but it will become an indispensable tool in the future. So, how should we university students face this?

As Professor Kawahara says, there is no other way than using generative AI such as ChatGPT. Rather, the ability to use it effectively will be needed in the future

. However, there is a risk in using AI to do all the core parts of the task. Currently, there are still quite a few mistakes in the generated AI's answers, and if you submit a report full of mistakes as is, your evaluation will naturally go down.

In other words, to make full use of generative AI, you need the ability to correct mistakes in answers

. To do this, make sure you understand what you are writing in your report. The image is that you can understand and write it yourself, but use AI as an aid to improve efficiency.

It is important to be aware that you are responsible for what you create

. I think teachers will have different opinions on whether or not it's okay to use AI for assignments, so this is just my personal opinion.

I think Professor Ogata is right.

It is important to use "critical thinking" to consider whether the answers provided by the generative AI are really correct

.

And we Faculty will also need to respond by changing the way we assign assignments and the way we evaluate them. After all, we are unable to cope with the rapid changes in generative AI.

By the way, do teachers also use generative AI?

I'm using it. I use it like a wall-based opponent, by getting them to come up with ideas and asking them what they think of the title of the paper.

Similarly, I have students come up with ideas and edit drafts of English papers into more sophisticated expressions.

What is the significance of universities in an era where the use of AI is commonplace?

As I listened to the teachers' stories, I realized that I wanted to be able to use generative AI skillfully. However, if AI becomes capable of doing more and more things, I start to wonder what the point of going to university is.

If you just ask a question and get an answer, it may be possible to do it with generative AI. However, at university, you can obtain cutting-edge knowledge from professors who are conducting research in various fields.

I think this is a very valuable experience.

In addition, all the generative AI answers is based on information available on the internet, and only knowledge that is already known. AI cannot teach us anything that no one else in the world knows yet, nor can it discover and solve new problems.

The most interesting thing about university is that new discoveries may be made through the pursuit of knowledge and research

.

Building human relationships is also an important function of university.

Connecting people is very meaningful, and university is a place where you can do that

.

...My current answer is pretty much the same as Watanabe-san's answer to ChatGPT about the "significance of university" (lol).

Lastly, could you give a message to Waseda University students?

The advent of generative AI had a really big impact. To be honest, we cannot predict how AI will evolve or what will happen to society in the future.

However, with the advent of generative AI, it has become clearer what only universities can do and what only humans can do

. People who happened to sit next to you in class, people who joined the same student club, people you did research with in the laboratory...I want you to cherish these encounters and make it a place where you can deepen your learning

.

President Aiji Tanaka has said that `` Intellectual resilience, Flexible sensitivity, and Responsive reasoning'' are important skills for students to acquire, and I believe this is exactly right.

With the advent of generative AI, even more outstanding intelligence will be needed

, so I would like to see it cultivated. Please cherish the power of discovery and invention and creativity that only humans can have

.

Personally, I had seen negative opinions about generative AI in the media, so I was hesitant to use it until today. However, listening to the teachers' stories made me want to think critically, use tools effectively, and refine my knowledge. Thank you for your time today!