AI Translation
I would like to address translation, or more specifically, AI translation programs such as Google and ChatGPT. While these tools are easy to access and less expensive, there are some issues that should be considered regarding them.
1. No interpretation or translation will ever be 100% accurate. You would have to have the same thought world as the person who is speaking or writing in order to know with 100% accuracy you were conveying all of their intent. A translator or interpreter will look at or listen to what is being conveyed, ask questions if they need to, refer to different parts of the document, and use their best judgment to translate or interpret what is written or what is said. They try very hard to maintain the same intent. Translating programs simply translates word for word as it appears.
2. No program, currently, is completely proficient in any language. Some of the most popular may be “pretty good”, but depending on what you are translating, is “pretty good” good enough? You must consider this: would it be ok if there were mistakes in your document? Mistakes could be grammatical, they could be wrong word choices, the intent could be conveyed incorrectly, or parts of the document may just not make sense. These issues are going to happen when you do not have a human involved.
3. Do you know the proficiency of the language in the program you are using? How would you know if what is translated in the target language (the language you are translating into) is correct? Having checks and balances is so important.
4. Is the translation localized? When you localize a translation, you are making sure it makes sense to the group of people you are translating it for. If the population in the area you are in is from a certain area of another country, there could be slang or words that are used in that area. This “local language” makes it easier for someone to understand what you are trying to convey. (Think of Pop and Soda here in America. Some areas say pop for carbonated sugary beverages, some say soda, and yet others use the brand name Coke as a general term for these same beverages).
DID YOU KNOW?
1. AI wants to be successful. If it needs to, it will create content necessary to finish a job. Example: There was an attorney who used AI to create his briefs for court. AI created a brief but also created court cases for the brief because it was unable to locate any cases to prove its points. A judge looked at the brief and tried to look up the court cases, only to find they did not exist. If the attorney had done it himself, he would have used accurate information. If the document had been proofread, this would also have been found before the judge read it.
2. AI learns from itself. When you submit a document to an open AI platform such as ChatGPT or Google, the document is on the open web essentially. The program will translate it, then use what it translated to learn. When it does that, it maintains the information from the document and is able to use it for other translations.
It is so tempting to use a free and seemingly easy program to do your translations, but when you consider the issues that may be involved, it is best to find a program that also has a human to proofread what you are having translated.
