The ChatGPT pricing plans are divided into four categories (Free, Plus, Team, Enterprise).

OpenAI has added a new ChatGPT "Team" plan to the existing Free and Plus plans. (January 10, 2024)

○ Features of the ChatGPT Team Plan

 1. Increased Usage Limits: Approximately 100 uses every 2-3 hours (Usage limit for Plus model: Approximately 40 uses every 3 hours) 

 2. Enhanced Security: OpenAI does not train on conversations/data shared by GPT.

 3. Easy Team Management: GPTs can be shared among members within the workspace. 

 4. Annual Subscription: $25 per month when subscribed annually (Monthly subscription: $30 per month)      

  * Payment is made for the number of initial seats when subscribing (additional seats can be added after initial setup, $30 per user) However, it is important to carefully consider adding users as there is no mention of refunding costs if users are reduced.

○ ChatGPT Pricing Plan Comparison

* Enterprise Plan (the highest-tier plan for corporate use) is priced after inquiries and consultations. 

( ※ Based on data as of March 25, 2024, functionalities may be added or fees may be changed.)

출처: 오픈AI 공식홈페이지(Pricing): https://openai.com/chatgpt/pricing

- 오픈AI 공식홈페이지(Team): https://openai.com/chatgpt/team

- https://spartacodingclub.kr/blog/chatgptplus

- https://contents.premium.naver.com/chatgpt/buff/contents/240207030551922xx

Here are the differences between ChatGPT, Bard, and BING AI:

(※ The information provided is based on the data available as of May 24, 2023. 

Additional features may have been added since then.)

Source: “챗GPT 장점 많지만 단점도 많아---빙GPT로 보완 가능하다”(Data&AI, 2023.03.21)

Bing AI 챗봇 vs Google 검색: 누가 더 나은가요, 그리고 광고는 어떤가요? (AdGuard, 2023.04.19)

[바드가 궁금해③] 바드와 챗GPT·빙 뭐가 더 나을까(뉴스핌, 2023.05.20)

[인사이드 스토리]인공지능 장착한 '빙', 챗GPT와 다른 것(하나 Bizwatch, 2023. 03.11)

챗GPT, 아이폰에서도 쓴다···모바일 앱 출시(경향신문, 2023.05.19)

[찬이의 IT교실] '챗GPT’ vs ‘바드’...검색 위한 AI 챗봇 비교(Ai타임스, 2023.05.18)

“수능 24번 풀어줘” 챗GPT-바드-빙 승자는?(중앙일보, 2023.05.23)

챗GPT 장점 많지만 단점도 많아(NSP 통신, 2023.03.21)

Here are some strategies to help you use ChatGPT more effectively:

1. Ask clear and specific questions

 - ChatGPT, unlike humans, may struggle with understanding context, so it's crucial to ask precise and detailed questions.

 - Instead of broadly framed questions, try to include the context and the specific information you're seeking.

(example) if you're asking about how to get to Busan:

Rather than asking, "How should I get to Busan?"

You could ask, "Given that I have a budget of 100,000 won, approximately three hours of travel time, and a starting point of Seoul, and considering that I'm uncomfortable with flying, what would be the best way for me to get to Busan?" This way, you're more likely to get the answer you're looking for.

2. Set expectations for the response upfront:

 - If you want to avoid potentially incorrect information, you can instruct ChatGPT by saying, "If you're unsure, please tell me that you don't know."

 - If you're dealing with complex issues or instructions, you can request a step-by-step response. 

 - Start your question with, "Please answer step-by-step." This strategy is particularly effective when you need to understand programming code sequentially.

3. Try asking questions in English:

 - While ChatGPT can handle various languages, its training data is primarily in English, hence it has a stronger understanding of English.

 - As a result, if you pose your queries in English, you can expect a higher level of accuracy and contextual comprehension compared to asking in Korean.

 Source: Chat GPT Utilization and Learning Ethics Video [Education Development Center (2023)]

Currently, OpenAI offers Chat GPT in three main versions: 

Legacy 3.5, Default 3.5, and 4.0. The differences between each version are as follows:

1. Legacy (GPT-3.5) 

The initial model of ChatGPT Plus, it is similar to Turbo mode but slower and generates longer responses.

2. Default (GPT-3.5) 

This version is in Turbo mode, with inference capabilities similar to the Legacy version but faster and generates more concise responses. It is suitable for immediately answering questions that do not require significant logical thinking.

3. GPT-4.0 

This is the most advanced and recent model, available only to Plus paid members. 

It has the highest inference capabilities and generates longer responses, but it is slower.

Programs that detect AI-generated text can help determine whether tools like ChatGPT or Bing AI have been used.

Notable AI detection programs include GPT Zero, Detect GPT, Originality.AI

with their respective features detailed below.

However, using a detection program cannot guarantee that the text was entirely created through AI generators like ChatGPT. Even after using a detection program, it is necessary for the instructor to verify the content again.

(Example) OpenAI, the developer of ChatGPT, correctly identified 26% of AI-generated text in a test evaluation using their AI text classifier as "potentially AI-generated," while wrongly labeling 9% of human-written text as "AI-generated."

1. GPT Zero

   1) Evaluation criteria: Perplexity score / Burstiness score-perplexity change

   2)AI-generated text tends to be consistently structured, unlike human-written text, because machines generate sentences according to programmed rules. AI also creates sentences by considering which word is most likely to follow another.

case 1  Judged as written by AI(9.857/2.476)

- Lower scores indicate an increased likelihood of being AI-generated

case 2  Judged as written by a human(53.750/44.262)

- Portions of the text with a likelihood of being AI-generated are highlighted in yellow

2. Detect GPT 

   1) Evaluation criteria: Not clearly specified

   2)After entering the content and clicking ’check content’ the program judges whether the content is AI-generated and provides the likelihood

3. Originality.AI

   1) Evaluation criteria: Uses linear probability models, "Burstiness," and "Perplexity" for evaluation

   2) Human-written text is highlighted in green, while AI-generated text is highlighted in red

Source: https://raj-k-kadiyala.medium.com/gptzero-vs-chatgpt-a-gray-story-901b825e0666

According to Article 7, Section 3 of the guidelines on handling attendance and examination misconduct, a student who uses AI-generated content like ChatGPT without the instructor's permission during an exam may be considered as engaging in cheating. 

[Guidelines on Handling Attendance and Examination Misconduct, Article 7, Section 3]

1. ChatGPT generates text by collecting information from online sources and does not provide citations. Using it as if you produced the content yourself may result in copyright infringement or violations of academic ethics (plagiarism). (학습윤리 가이드, 연구윤리정보센터)

2.The copyright status of AI-generated content is unclear, but the ChatGPT terms of use include requirements and restrictions regarding the use of output. ChatGPT users are responsible for the use of generated content and must ensure it does not violate any copyright laws.

3. Unlike ChatGPT, BING chatGPT provides citations for the sources of information. Using it in conjunction with ChatGPT and including the source information can help avoid violations of academic ethics.

Source: "chatGPT - 법의 회색 영역: 저작권과 표절의 문제" (The Science Plus, 2023.03.18.)

              "chatGPT, 어디까지 허용?...새학기 대학가 뜨거운 논쟁" ( 한경, 2023.03.13)

The following points should be kept in mind when using ChatGPT and other AI tools, as they cannot fully replace human thoughts or writings. After reviewing the text generated by ChatGPT, it's essential to carefully examine and edit the content based on your knowledge and experience. Other precautions include:

1. Clear and specific questions are better than abstract ones.

For example, instead of asking, "How's the weather?" it's better to ask, "What's the weather like today in Jongno District, Seoul?" This will yield more specific and effective answers.

2. Verify the facts in the generated text before using the output.

ChatGPT and other AI tools are based on large language models (LLMs). They are trained on massive amounts of data obtained from various texts, such as books, blogs, articles, and the internet. Therefore, they may not be accurate when creating content on specialized or unpopular topics. 

3. ChatGPT generates answers based on a large amount of online data and probabilities, which may include distorted facts.

ChatGPT does not disclose the sources of information in ways like links or citations. Since its data includes text from the internet, the answers may reflect biases, stereotypes, or minority opinions found online. To address this issue, you can cross-check with other GPT resources like Bing AI to increase accuracy.

Source: " "ChatGPT, if you can't stop it, use it correctly"" (kunews, 2023.03.21.)

1. Please evaluate the process-oriented evaluation. (Presentation to perform the task step by step)

The process-based evaluation checks the progress step by step and produces the final result,

You can avoid problems such as plagiarism or lack of fact-checking due to using Chat GPT when doing your final assignment without sufficient planning.

2. Please use the open book test, oral evaluation, and critical evaluation during class.

It is not reflected in grades such as peer review, self-evaluation, and mutual evaluation, but using critical evaluation methods that can improve the learning process and results can reduce the possibility of plagiarism by allowing students to review assignments.

3. Please evaluate to utilize multidimensional thinking capabilities such as critical thinking.

Students can use an evaluation method that presents problems that analyze and evaluate the validity or lack of Chat GPT answers based on the information collected through Chat GPT. (ex. Make computer students discover defects in the code formed by Chat GPT)

Source: "고려대, 국내 대학 최초 chatGPT 활용 가이드라인 제정" (boann뉴스, 2023.03.17.)

GPT Zero, Detect GPT, Originity as an AI Detect Program that detects text written by AI.There is AI and so on.

Detectors can learn subtle differences and patterns between the two types of text based on data containing both human-written and AI-generated text to predict whether a given text was created by artificial intelligence or human-generated.

However, as AI-generated text continues to evolve, it is becoming increasingly difficult to accurately identify AI-generated text. The instructor needs to check and judge the results of using the detection program once again.

For example, the AI text classifier released by OpenAI, which developed Chat GPT, correctly identified 26% of AI-written text as 'likely AI-written' in the trial evaluation, and incorrectly marked 9% of human-written text as 'AI-written text'.

Source: "AI 대폭발' 불러올 챗GPT…저작권 표절 개념 재정립 필요성 커져"(서울경제, 2023.3.31)

            "표절 안걸린다?...논문 쓰는 AI, 과학계 '윤리 문제 우려'" (Futurism뉴스, 2023.01.19.)