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, the Centre for Teaching, Learning and Technology (CTLT) at UBC has performed tool evaluations for on-device usage of gpt4all and Ollama. CTLT is also offering a session regarding the intersection of Open Educational Resources (OER) and AI. OpenAI released an educational framework called Swarm, designed to coordinate and delegate tasks to multiple agents. Rhymes AI released “the world’s first open-source, multimodal native Mixture-of-Experts (MoE) model”, Aria. Another open-source video generation tool named Pyramid Flow was introduced by a collaboration of researchers from Peking University, Beijing University of Posts and Telecommunications, and Kuaishou Technology, the creators of Kling AI. Amazon has implemented AI into their vans, allowing delivery drivers to locate packages more quickly. Google has updated its image generation tool, Imagen v3, to provide better text rendering and higher-quality images. Zoom has introduced Zoom Avatars to speak in Zoom calls on behalf of the user, and finally, Mistral AI has introduced new AI models that are optimized for consumer-grade devices such as laptops and phones.
Here’s this week’s news:
CTLT’s Locally Hosted Tool Evaluations
UBC’s Centre for Teaching, Learning, and Technology has released tool evaluations for the on-device usage of gpt4all and Ollama. These tools are used for running large language models on local devices, and the evaluation aims at examining security, versatility, and capabilities of models that these platforms provide. The purpose of these tools is to provide guidelines for responsible AI use and help instructors make an informed decision about using AI models in a local environment.
Check out the full evaluations here.
OER and GenAI – October 23, 2024
The “OER and GenAI” event on October 23, 2024, hosted by UBC’s Centre for Teaching, Learning, and Technology, will explore the intersection of Open Educational Resources (OER) and Generative AI. It aims to discuss how AI tools can support OER development and enhance teaching and learning practices. The session invites educators to examine the opportunities and challenges, as well as the ethical considerations that AI technologies bring to the creation and use of open educational resources.
OpenAI’s Swarm Framework
OpenAI’s Swarm framework is an educational tool for orchestrating multi-agent systems. It enables developers to create agents that coordinate to solve complex tasks by dividing work among specialized agents. Swarm is designed to run client-side and can execute tasks like function calling, context management, and agent switching. It supports lightweight, multi-agent workflows without persistent state, making it ideal for exploring agent-based AI solutions.
Check out the Github repository for Swarm.
Rhymes AI releases Aria
Rhymes AI has introduced Aria, the first open-source multimodal native Mixture-of-Experts (MoE) model. Aria excels in handling diverse input types like text, images, and video by activating only relevant parts of the model, reducing memory use. It performs well in tasks such as long video understanding and math problem-solving, marking a potential shift towards efficient multimodal AI tools for consumers.
Pyramid Flow Open Source AI Video Generation
Pyramid Flow, a new AI-powered video generator, has been launched as a fully open-source tool. This model delivers high-quality video generation while being accessible on consumer-level devices. Pyramid Flow’s open-source nature allows developers and content creators to integrate and customize video generation capabilities without the need for expensive hardware, making it a significant step forward in democratizing AI-driven content creation.
Amazon VAPR
Amazon has introduced an AI-powered technology into their delivery vans, called the Vision-Assisted Package Retrieval (VAPR). This tool highlights the correct packages with a green light, allowing delivery drivers to quickly locate correct packages, and save time on searching through hundreds of packages that are to be delivered each day. Amazon plans to deploy 1000 electric powered delivery vans with VAPR installed by early 2025.
Google Gemini Imagen v3
Google has updated its Gemini image generation tool, Imagen v3, now allowing for the creation of higher-quality images. Imagen v3 highlights better text rendering and greater versatility with prompt understanding as improvements from their previous model. While the model stays away from generating faces of known individuals for ethical concerns, the paid version can handle more detailed tasks, including people and faces, offering competitive options for visual content creation.
Check out Google’s Imagen v3 here.
Zoom’s AI Avatars
Zoom has launched new AI-powered features, including AI avatars that represent users in meetings and the ability to generate short video clips that can speak on behalf of you during calls. These avatars aim to provide more dynamic meeting experiences by using animated characters, while the clips feature allows users to pre-record messages, reducing the need for live participation. These innovations are part of Zoom’s broader effort to enhance remote work interactions.
Mistral’s New AI models for Edge Devices
Mistral has released new AI models optimized for edge devices, making it possible to run powerful machine learning models in smaller, consumer-grade hardware. The family of models called “Les Ministraux” includes two models, Ministral 3B and Ministral 8B, both boasting a context window of 128,000 tokens. This opens the door for AI applications that perform tasks rather than just generate text, offering more functional capabilities on mobile and other lightweight devices.
Tool of the Week
Tool of the Week: Google Illuminate
What is Google Illuminate?
Google Illuminate is an experimental AI tool designed to adapt academic content to user learning preferences by generating audio discussions of research papers. It converts academic papers from arxiv.org into podcast-styled conversations, highlighting key points and making complex information more accessible. It is currently optimized towards papers in the domain of computer science.
How is it used?
Users select papers by searching or pasting arxiv.org links, then Illuminate generates audio conversations summarizing key points. Users can also customize settings to tweak the intended audience level, duration, and tone of the discussion. The platform allows for up to five audio generations per day.
What is it used for?
Illuminate helps users explore and understand academic papers through engaging audio summaries, enhancing the research experience and allowing for an alternative method to approach learning from academic papers.
For more information, test out Google Illuminate here.
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 […]