Chat GPT vs. Humans. The text made by Chat GPT makes it impossible to separate the text, made by humans.
Chat GPT vs. Humans. The text made by Chat GPT makes it impossible to separate the text, made by humans.
Text that Chats GPT made is impossible to separate from human-made texts. This thing is problematic in the case of plagiarism. Another problematic thing is the thing, that even the best scanners cannot separate texts that Chat GPT made from texts that humans made. The fact is that cheaters might use the AI rather at home than in an office or they will copy-paste the text that Chats GPT made into an empty Word document. Or they can simply write the text that they see on the cell phone's screen into the documents.
So even if the company denies access to the Chat GPT or Bing homepages, that will not affect things like cheating with the Chat GPT or Bing. AI is a rising tool, that has shown their potentials. And this kind of thing is always problematic because the company's purpose is to make money for its owners.
"Linguistics experts struggled to differentiate between AI- and human-generated writing, with a positive identification rate of only 38.9%, according to a new study. Despite logical reasoning behind their choices, they were frequently incorrect, suggesting that short AI-generated texts can be as competent as human writing." (ScitechDaily.com/ChatGPT vs. Humans: Even Linguistic Experts Can’t Tell Who Wrote What)
And in that case, there is needed an effective attitude for jobs. AI is a tool that makes work faster and more effective than people. AI is a tool that companies must learn to use. The work's purpose is to make money for people. And people must do all the work effectively. The data systems' purpose is not to make things more difficult. Its mission is to make life easier, not to make things harder than they are.
But the problem is that AI doesn't ever think about things, that it writes. This is the reason, why people should always check sources by going to those home pages. And in that case, they check the texts and their background. If there is some homepage where there is fake information. The AI will not recognize the data itself, or what the homepage involves. And that causes situations where the AI can make mistakes.
The mistake is in the home pages that the AI selects. And that is the reason why people who use AI must have enough skills to make the estimation. People can improve the AI's skills to give it trusted information is to preselect the homepages list there is confirmed. The user can make a list of the links to the valid home pages. That helps the AI to select trusted information.
Things like Scientific America are trusted sources, so the users should have the ability to determine what sources the AI should use. That thing decreases the possibility of making errors, because of the contents of the homepages. The fact is, that AI is one of the most powerful tools. That means that kind of tool is extremely important for companies. The AI is the tool that makes work easier. And that makes companies give training for people to use that kind of system.
https://scitechdaily.com/chatgpt-vs-humans-even-linguistic-experts-cant-tell-who-wrote-what/
Comments
Post a Comment