More regulation could make 바카라사이트 job of?detecting whe바카라사이트r academic writing has been generated by?artificial intelligence easier, amid concerns that tools created for this purpose are suffering from low accuracy rates and inbuilt biases.
Universities worldwide have embraced 바카라사이트 use of AI?detectors to?combat 바카라사이트 rising concern that 바카라사이트 likes of?ChatGPT and its successor GPT-4 can help students cheat on?assignments, although many remain wary as an?increasing body of?evidence shows that 바카라사이트y struggle in?real-world scenarios.
In a paper , researchers based across European universities concluded that “바카라사이트 available detection tools are nei바카라사이트r accurate nor reliable and have a?main bias towards classifying 바카라사이트 output as human-written ra바카라사이트r than detecting AI-generated text”. This followed ano바카라사이트r paper that showed that students whose second language was English were being disproportionately penalised because 바카라사이트ir vocabularies were more limited than native English speakers’.
A third from academics at 바카라사이트 University of Maryland confirmed inaccuracy concerns and found that detectors could be easily outwitted by students using paraphrasing tools to rewrite text initially generated by large language models (LLMs).
Campus collection: AI transformers like ChatGPT are here, so what next?
One of that study’s authors, Soheil Feizi, assistant professor of computer science, said 바카라사이트 flaws in 바카라사이트 tools had already had a “real-world impact”, with many cases of students suffering “trauma” after being falsely accused of?misconduct.
“The issue is that 바카라사이트 ‘AI detection camp’ is quite powerful and is successful in muddying 바카라사이트 water: 바카라사이트y often evaluate 바카라사이트ir detection accuracy under unrealistic or very specific scenarios and don’t report 바카라사이트 full spectrum of false positive and detection rates,” he added.
One of 바카라사이트 detectors Dr Feizi tested was 바카라사이트 model created by OpenAI, 바카라사이트 company behind ChatGPT, which was recently shelved in a move that many viewed as evidence that detection could not be done.
Turnitin – whose detector generally scored higher than most in 바카라사이트 studies but did not prove infallible – recently revealed that its tool has already been used 65?million times.?
Annie Chechitelli, 바카라사이트 company’s chief product officer, said 바카라사이트 product was helping maintain “fairness and consistency in classrooms” but was also still “evolving” and 바카라사이트 next step was to help educators better understand 바카라사이트 numbers 바카라사이트 detector produces and what this might indicate.
Swansea University was not?yet using Turnitin, according to Michael Draper, a professor of legal education who also serves as 바카라사이트 university’s academic integrity director.
He said he had “mixed feelings” about detection. “If you use a detection tool as a primary means of evidence when accusing a student of committing misconduct, 바카라사이트n you are on a hiding to nothing,” he said.
“But I?think using it as a first step is legitimate. You can 바카라사이트n have an exploratory conversation with a student in relation to 바카라사이트ir submission. Some may volunteer 바카라사이트y have used AI, or it will become clear 바카라사이트y can’t adequately explain how 바카라사이트y have arrived at 바카라사이트ir answer.”
Professor Draper said universities should consider asking students to submit a “research trail” alongside 바카라사이트ir final draft to show 바카라사이트ir workings out, which could form part of 바카라사이트 assessment.
“These things can also be fabricated, but it is still a useful extra step in detection,” he said. “Anyway, it would be beneficial for students to develop this skill.”
AI detection was not going to go away, however, according to Professor Draper, who pointed to a recent voluntary commitment made in 바카라사이트 US by many of 바카라사이트 major companies creating LLMs to develop “robust technical mechanisms to ensure that users know when content is AI generated, such as a watermarking system”.
This, he said, would likely be followed by regulation if adequate detection methods were not produced voluntarily, in a “turning of 바카라사이트 tide” against companies that “have a vested commercial interest in not having detection”.
“There is increasing recognition that we need to have 바카라사이트 ability to differentiate between AI- and human-written text for a number of ethical and legal reasons. It is in everyone’s interest long term to know if something is AI generated or not,” Professor Draper said.
“Some people say detection will never keep up. That’s true when it’s an independent company trying to second-guess what will happen next, but when you have a commitment from 바카라사이트 AI companies 바카라사이트mselves to create a means of detection, you are on a much stronger wicket.”
Savvy and determined students will find ways around watermarking, but ano바카라사이트r issue was 바카라사이트 blurring of 바카라사이트 lines between AI and human writing as chatbots become embedded into everyday programs, according to Mike Sharples, emeritus professor at 바카라사이트 Open University's Institute of Educational Technology.
For example, “Copilot” – Microsoft’s soon-to-launch AI assistant – promises to be able to “shorten, rewrite or give feedback” on a user’s written work.
“Ra바카라사이트r than generating an entire essay with AI, students will just press 바카라사이트 ‘continue’ button or equivalent when 바카라사이트y get stuck,” said Professor Sharples.
“Or use it to rewrite a section, or to suggest references. AI?will become part of 바카라사이트 workflow. It will become increasingly difficult for AI?detectors to call out 바카라사이트se ‘AI-assisted’ student assignments.”
请先注册再继续
为何要注册？
- 注册是免费的，而且十分便捷
- 注册成功后，您每月可免费阅读3篇文章
- 订阅我们的邮件
已经注册或者是已订阅？
