GenAI Studio: News, Tools, and Teaching & Learning FAQs
These sixty minute, weekly sessions – facilitated by Technologists and Pedagogy Experts from the CTLT – are designed for faculty and staff at UBC who are using, or thinking about using, Generative AI tools as part of their teaching, researching, or daily work. Each week we discuss the news of the week, highlight a specific tool for use within teaching and learning, and then hold a question and answer session for attendees.
They run on Zoom every Wednesday from 1pm – 2pm and you can register for upcoming events on the CTLT Events Website.
News of the Week
Each week we discuss several new items that happened in the Generative AI space over the past 7 days. There’s usually a flood of new AI-adjacent news every week – as this industry is moving so fast – so we highlight news articles which are relevant to the UBC community.
In this week’s tech news, Sam Altman claims AI will transform society and lead to unprecedented advancements in various fields. OpenAI introduces Advanced Voice Mode, enabling advanced conversational capabilities for its models. Anthropic launches a contextual retrieval system that enhances AI’s ability to retrieve relevant data. Microsoft has signed a deal to purchase nuclear energy to power its data centers, and has unveiled “Correction,” a tool aimed at reducing AI hallucinations. LinkedIn is found to have used user data to train AI models before updating its privacy policy, raising concerns about consent and data usage. Google’s GNoME AI has discovered millions of potential new materials, which may lead to breakthroughs in technology like batteries and superconductors. The University of Notre Dame is hosting a panel discussing the impact of AI on teaching methods, and UBC is hosting a session providing strategies for educators to develop assignments using generative AI. The Allen Institute has launched Molmo, a powerful open-source multimodal AI model that can process both text and images, while Meta released Llama 3.2, a new series of AI models designed for mobile and edge applications.
Here’s this week’s news:
CEO of OpenAI on the “Intelligence Age”
Sam Altman, the CEO of OpenAI, envisions an “Intelligence Age” driven by AI advancements, leading to unprecedented societal progress. He predicts AI will enable humans to solve complex global challenges and enhance everyday life, such as personalized education and healthcare. However, managing risks, like labor market shifts, will be crucial. Read more.
OpenAI introduces Advanced Voice Mode
OpenAI has introduced Advanced Voice Mode, which offers an engaging and natural way to converse with ChatGPT. This feature can be used in more than one language in conversation. Watch the full video.
Anthropic’s Contextual Retrieval AI
Anthropic has launched a new approach called “Contextual Retrieval” to enhance AI’s ability to fetch relevant information based on nuanced user requests. This technology could lead to more reliable and context-aware responses, especially in complex or ambiguous scenarios. Read more here.
Three Mile Island to Power Microsoft Data Centers
Constellation Energy plans to restart the Three Mile Island nuclear plant’s Unit 1 reactor under a 20-year deal with Microsoft to provide carbon-free energy for its data centers. This initiative supports Microsoft’s goal to become carbon-negative by 2030 and reflects renewed interest in nuclear power to meet growing electricity demands from AI and cloud services. Read more here.
Microsoft’s New Tool to Address AI Hallucinations
Microsoft introduced “Correction,” a tool aimed at reducing AI hallucinations by automatically fact-checking AI-generated text against trusted sources. Available through the Azure AI Content Safety API, this tool seeks to enhance the reliability of generative AI. However, experts remain cautious, noting that while Correction may improve accuracy, it won’t eliminate the root causes of hallucinations in AI systems. Read more.
LinkedIn Scraped User Data for AI Training
LinkedIn has been found to have scraped user data for AI model training before updating its terms of service to reflect this. The move raises privacy concerns of the opt-out model and highlights the tension between users and social media companies due to the increasing number of platforms using user-generated content to train AI models. Read more here.
Google’s GNoME AI Accelerates Material Discovery
Google’s DeepMind team has developed GNoME (Graph Networks for Materials Exploration), an AI system that has discovered 2 million new chemical compositions of inorganic crystals that could be potentially synthesized. These discoveries, which could take centuries using traditional research methods, may lead to advancements in technologies such as batteries and superconductors. Read more here.
University of Notre Dame Faculty Panel: Teaching in the Age of AI – September 27, 2024
The University of Notre Dame is hosting a faculty panel discussing the role of AI in teaching. The session will explore how educators can integrate AI responsibly into their educational practices and their classrooms. Learn more.
UBC GenAI Maker Session: Developing Quizzes and Problem Sets – October 8, 2024
UBC is hosting an online session to guide educators on creating quizzes and problem sets using generative AI. This session will focus on practical strategies and provide hands-on time for participants to develop their materials in a collaborative setting. Register here.
Molmo: Open-Source Multimodal AI
The Allen Institute for AI (Ai2) has released Molmo, an open-source multimodal AI model that can interpret images and process tasks through a chat interface. Designed to power AI agents, it allows developers to create tools capable of handling complex computer tasks like navigating file directories or drafting documents. Molmo’s fully open-source nature sets it apart from commercial models, providing extensive customization options for developers and making it accessible for startups and researchers alike. Read more.
Meta releases Llama 3.2 – Lightweight and Multimodal Llama Models
Meta has released Llama 3.2, a new version of its open-source AI models, which includes a variety of different sized models (1B to 90B parameters). These models are optimized for edge and mobile devices, with the larger models supporting advanced image understanding and smaller models supporting multilingual text generation. Llama 3.2 offers flexibility with pre-trained and instruction-tuned versions, enabling developers to build privacy-focused applications that run locally. Read more.
Tool of the Week
Tool of the Week: NotebookLM
What is NotebookLM?
NotebookLM is Google’s AI-powered assistant, designed to help users better manage, organize, and understand large volumes of information. By taking in data from a variety of sources NotebookLM acts as a personal research assistant, allowing users to query their notes, generate summaries, and find relevant insights quickly.
How is it used?
Users can upload documents from their device or Google Drive, as well as add links to other websites, and then ask questions or generate summaries from the content in the sources. NotebookLM also helps by automatically identifying key topics or extracting important points from large bodies of text. This makes it particularly useful for research, academic writing, or project management, where users deal with extensive information.
What is it used for?
NotebookLM is ideal for students, researchers, and professionals who need to navigate complex topics or large datasets. It streamlines workflows by offering AI-driven insights and summaries, helping users focus on high-level concepts while reducing the time spent manually sifting through content. Its ability to answer specific questions based on document content makes it an invaluable tool for productivity and learning.
For more information, explore NotebookLM.
Without a PIA, instructors cannot require students use the tool or service without providing alternatives that do not require use of student private information
Questions and Answers
Each studio ends with a question and answer session whereby attendees can ask questions of the pedagogy experts and technologists who facilitate the sessions. We have published a full FAQ section on this site. If you have other questions about GenAI usage, please get in touch.
-
Assessment Design using Generative AI
Generative AI is reshaping assessment design, requiring faculty to adapt assignments to maintain academic integrity. The GENAI Assessment Scale guides AI use in coursework, from study aids to full collaboration, helping educators create assessments that balance AI integration with skill development, fostering critical thinking and fairness in learning.
-
How can I use GenAI in my course?
In education, the integration of GenAI offers a multitude of applications within your courses. Presented is a detailed table categorizing various use cases, outlining the specific roles they play, their pedagogical benefits, and potential risks associated with their implementation. A Complete Breakdown of each use case and the original image can be found here. At […]