🗞 News of the day

🤖 Amazon AI Chatbot

Amazon is developing a new AI chatbot

• Metis Chatbot Development: Amazon is working on a new chatbot called Metis, powered by an AI model named Olympus. Unlike previous models, Metis uses a retrieval-augmented generation (RAG) approach, which allows it to access up-to-date information from external sources, leading to more accurate and clear responses.

• Key Features and Control: Metis' RAG approach redirects the language model to retrieve information from authoritative sources, giving organizations better control over the generated text. Users gain insights into how responses are generated, enhancing transparency and trust in the chatbot's answers.

• Team and Competition: Amazon's CEO Andy Jassy and head scientist Rohit Prasad are leading the development, with a team of veteran Alexa AI workers. Despite facing competition from OpenAI's ChatGPT and Google's Bard, Amazon plans to release Metis in September, aiming to make a significant impact in the AI industry with its unique capabilities.

🔖 Today's Cool Story

🚨 AI Outperforms University Students

AI Outperforms University Students in Exams, Study Shows

Researchers at the University of Reading created 33 fake students and used ChatGPT to answer psychology exams. The AI students scored, on average, half a grade boundary higher than actual students. Even more shocking, 94% of these AI-generated essays didn't raise any red flags with the exam markers.

The study, published in Plos One, suggests that students could easily use AI to cheat and end up with better grades. The researchers behind the study, Associate Prof Peter Scarfe and Prof Etienne Roesch, are calling this a huge wake-up call for educators everywhere.

They warn that the education sector needs to evolve quickly to keep up with AI's growing influence. Dr. Scarfe even mentioned that while we might not go back to handwritten exams, something's gotta give in the way we assess students.

Interestingly, the AI outperformed real students in first- and second-year exams but fell short in third-year exams. This might be because AI struggles with more abstract reasoning, something third-year exams often require.

The study's findings have already led some universities, like Glasgow, to bring back in-person exams for certain courses. It's clear that the rise of AI in education is shaking things up, and schools will need to rethink how they handle assessments to keep things fair.

🚀 Try New Education Prompts

 🎨 Image Prompts

Prompt #1: Introspective brooding young man

Prompt #2: Storybook-inspired village

Prompt #3: Sculptural rock formations

See all Midjourney prompts →