Alona Fyshe, an associate professor at the University of Alberta and a fellow with the Alberta Machine Intelligence Institute , about OpenAI’s ChatGPT program.Sign up to receive daily headline news from the Calgary Herald, a division of Postmedia Network Inc.By clicking on the sign up button you consent to receive the above newsletter from Postmedia Network Inc. You may unsubscribe any time by clicking on the unsubscribe link at the bottom of our emails or any newsletter. Postmedia Network Inc.
How do you train something like this? Are you teaching it how to have conversations by having it read other conversations?Yes, so they trained it using rounds of back-and-forth conversations between people and then they would have new people write what the next turn of the conversation would be. And then they trained the model to try and generate the next turns.
What about the people who would just use ChatGPT to do their homework or write a paper for them; is there a way to curb that?: I think a more forward-looking thing to do would be to ask, “How can we change our assignments so that ChatGPT is not useful?” Or say, “Go ahead and use ChatGPT but the assignment is not testing the things that ChatGPT can do well, we’re testing something else.
I think the more alarming thing than generating your assignments is generating misinformation, getting ChatGPT to write an article about a conspiracy theory that could provide a takedown of a politician. We’ve had problems with external forces on social media trying to change public opinion and I think ChatGPT or systems like it could make that easier.Share this article in your social network
Education Education Latest News, Education Education Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Source: GlobalNational - 🏆 81. / 51 Read more »