#models

#models

This AI Can Do Schoolwork Well, if you’re looking for a machine that can easily churn out an essay, of course. Tech company OpenAI released ChatGPT, an AI chatbot that can easily generate school essays for any grade level.The software’s language model can also answer open-ended analytical questions, and write different kinds of literature such as jokes, poems, stories, and even a script for a YouTube video. The developers of this AI also made sure that it can create computer code. This machine is now a one-stop shop for homework for both lower-level and tertiary-level assignments, yikes.The rise of this development in AI can be detrimental to school curriculums, according to writing coach and founder of Crush the College Essay Peter Laffin. "The introduction of new artificial intelligence technologies into schools that enables students to auto-generate essays has the capacity to blow up our entire writing education curriculum," he shared with Fox News. While it can force teachers to change up their syllabi, students having easy access to tools like ChatGPT can just submit generated essays without internalizing what they’ve learned from class. The sad reality is, with an AI as good as the previously mentioned chatbot, teachers won’t be able to spot the difference. "The more easily available this is for younger students, the more problems this will create," Laffin added.Image credit: OpenAI #artificialintelligence #chatbot #ChatGPT #models #languagemodels #school #schoolwork 

Webcam Peeking Attack: Researchers Can Read People's Screens During Zoom Meetings by Reading Reflections in Their EyeglassesWell, this seems to be a new cause for concern for those who find themselves frequently in online meetings.Researchers have found a threat called “webcam peeking attack,” which involves a program that can reconstruct text reflected in an online meeting participant’s eyeglasses or other reflective objects in a video conference. This kind of threat may be used in the future if webcam technology is be further improved and developed. The academics involved in the study created a model of how this attack will work and it was found to have an accuracy of over 75% when reconstructing and recognizing text obtained by a 720p webcam. If this threat appears in the future, attackers could also identify the websites and other forms of information their victims are using. It can certainly be a tool for phishing important and sensitive information. Yikes! Image credit: Diva Plavalaguna/Pexels#cybersecurity #threat #research #technology #webcams #onlinemeetings #webcampeekingattack #textrecognition #models 