ChatGPT is useful as an interactive AI tool in many situations, but did you know that certain words and expressions are restricted when using it? There are reasons for this restriction to ensure safe and ethical use of AI.
In this article, we will explain in detail the types and background of words that cannot be used in ChatGPT, as well as workarounds when you encounter such words. In order to maximize the potential of AI, let's understand how to use it correctly.
table of contents
Types of words that cannot be used in ChatGPT
ChatGPT restricts the use of certain words and expressions for ethical and safe use. It can be broadly divided into the following four categories.
Words related to discrimination and prejudice
Discriminatory or offensive expressions regarding race, ethnicity, gender, religion, sexual orientation, disability, etc. are prohibited. This is extremely important in order to maintain an environment where everyone can use ChatGPT with peace of mind.
【example】Derogatory terms towards specific races or ethnic groups, sexist remarks, words ridiculing people with disabilities, etc.
Words related to violence and crime
You may not use expressions that encourage violence or crime, such as murder, terrorism, abuse, or self-harm. In order to realize a safe society, ChatGPT strictly restricts these expressions.
【example】Instructions on how to murder, how to make bombs, how to harm others, etc.
Obscene/inappropriate language
Sexually explicit expressions, obscene language, and expressions that violate public order and morals are also prohibited. To promote healthy communication, these words should be used sparingly.
【example】Content related to sexually explicit depictions, obscene slang, child pornography, etc.
Terminology related to security and privacy
Generating, sharing, or attempting to illegally obtain personal information (name, address, phone number, email address, credit card information, etc.), confidential information (passwords, login information, financial information, etc.) is prohibited. This is an important measure to protect your privacy and safety.
OpenAI's terms of service clearly state that personally identifiable information and private information should not be shared.It has been. This includes not only information about yourself, but also information about others.
【example】Personal information of specific individuals, confidential company information, hacking methods, etc.
If you would like to know more about specific prohibited words and the risks of using prohibited words, please check the article below.
"Specific examples of ChatGPT's prohibited words and the risks of their use”
Why are these words banned?
ChatGPT is an AI that learns from vast amounts of data and generates human-like text. if,If discriminatory and violent expressions are incorporated as learning data, there is a possibility that they will be reproduced.There is. This can cause major social problems, so by setting banned words, we minimize this risk.
In addition, protection of personal and confidential information and respect for copyrights are also cited as reasons for the prohibition.ChatGPT aims to provide ethically-free information.
If you repeatedly violate the Terms of Use, you may be subject to penalties such as account suspension or access restrictions. Please always check the latest terms of use and use it appropriately. For detailsOpenAI Usage PolicyPlease take a look.
Workarounds when encountering prohibited words
If you encounter a prohibited word in ChatGPT and can't get the output you want, try the following methods.
use paraphrase expressions
Avoid direct expressions and use more indirect expressions. For example, instead of using the word "kill" you can use words like "eliminate" or "remove".
【example】“Defeat the enemy” → “Suppress the enemy” “Overcome the problem”
Explain the situation more specifically
Describing the situation in concrete terms, rather than in abstract terms, helps ChatGPT generate appropriate answers.
【example】“Tell me a good method” → “Tell me an effective method to solve the problem 〇〇”
Clarify the intent of the question
By clearly communicating what you want to know, ChatGPT will be able to provide information without misunderstanding.
【example】“About 〇〇” → “Please tell me about the advantages and disadvantages of 〇〇”
Rethink your purpose
Let's reconsider whether we really need to obtain that information through ChatGPT. You should refrain from using it if it may lead to illegal or ethically questionable activities.
summary
Having a better understanding of ChatGPT's prohibited words will enable safe and reliable communication. You can bring out the performance of ChatGPT more effectively by not only keeping in mind the wording that follows the terms of use, but also by incorporating flexible paraphrases and ideas for prohibited expressions. Let's make the most of the possibilities offered by AI while creating an environment where it can be used with peace of mind.