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, Arizona State University has partnered with OpenAI to explore impactful uses of ChatGPT Enterprise in research and higher education. The University of Alberta introduced the “Artificial Intelligence Everywhere” certificate program designed to provide students with foundational knowledge in AI applications across various fields. The University of British Columbia’s Teaching and Learning Enhancement Fund has approved funding for several AI-driven proposals in 2025. Sakana’s AI Scientist-v2 generated a research paper that successfully passed peer review at a workshop during a machine learning conference. UBC sent out a message to the UBC community advising against downloading and using the DeepSeek Applications due to data security and privacy risks.
QwenLM introduced Qwen-Q 32B, a 32-billion parameter language model with a focus on STEM inquiries that can be run on personal computers released in recent years. Google released Gemma 3, another open-source AI model that performs well compared to larger models. The Allen Institute for AI launched Olmo2-32B, the first fully open-source model proven to match or outperform Open AI’s GPT-3.5 Turbo and GPT-4o mini in a variety of commonly-used benchmarks. Mistral AI announced Mistral Small 3.1, a 24B parameter model that outperforms similar models such as Claude-3.5 Haiku and GPT 4-o mini in benchmarks such as multilingual processing, text instruct, and multimodal instruct. Cohere introduced its most recent model, Command A, a compute-efficient yet high-performing model that can run on smaller devices.
Sesame conducted research on crossing the “uncanny valley” of AI conversational voices with the goal of developing digital voice assistants with natural, expressive speech. Butterfly Effect released Manus, a general AI agent that uses multiple AI models and operating agents to complete tasks. Jenni AI is an AI-powered research assistant designed to assist users in reading, writing, and organizing research materials. Google unveiled new features and upgrades for no additional cost. Finally, a recent audit reveals that western AI chatbots are unknowingly repeating Russian disinformation in their responses.
The Centre for Teaching, Learning and Technology (CTLT) at UBC is hosting several GenAI-related sessions in April. On April 9, 2025, the CTLT will host a session to discuss designing courses using AI. On April 16, 2025, there will be a session about developing AI literacy through interactive activities. On April 23, 2025, there will be a session about using AI to develop teaching materials. Finally, on April 30, 2025, there will be a session about using GenAI to design and build assessments.
Here’s this week’s news:
ASU Collaborates with OpenAI to Integrate ChatGPT Enterprise
Arizona State University (ASU) has partnered with OpenAI to integrate ChatGPT Enterprise across various disciplines. ChatGPT Enterprise differs from the open ChatGPT version in that it is designed for use by businesses and enterprises by incorporating more security and customizable features. By doing so, leaders at ASU position the institution as a leader in utilizing AI tools in higher education, research, and business. Read more.
University of Alberta Introduces “Artificial Intelligence Everywhere” Certificate
The University of Alberta has launched the “Artificial Intelligence Everywhere” certificate program. The program is designed to provide undergraduate students with foundational knowledge in AI applications across various fields. Students will learn basic concepts about AI, computer programming, and machine learning, thus preparing them to navigate an increasingly AI-driven world. Read more.
UBC Funds Innovative AI Projects Through TLEF
The University of British Columbia’s Teaching and Learning Enhancement Fund (TLEF) has approved funding for several AI-driven proposals in 2025. These funded proposals show that UBC is headed towards further integrating AI into teaching and research spaces across a diverse array of disciplines. Read more.
Sakana’s AI Scientist Achieves Peer-Reviewed Publication
Sakana’s AI Scientist-v2 successfully generated a research paper that passed the peer-review process at a workshop at a machine learning conference. The AI Scientist autonomously formulated hypotheses, conducted experiments, and authored three manuscripts without human modifications. Peer-reviewers at the workshop were notified that three of the forty-three papers submitted would be AI-generated, but they were not informed of which three papers would be AI-generated. One AI-generated paper crossed the acceptance threshold with a higher score than many of the human-written submissions. Read more.
UBC’s Restricts the Use of DeepSeek AI at UBC
UBC sent out a message earlier this week informing the UBC community of DeepSeek’s privacy and security risks. The article strongly advises members of the UBC community to discontinue their use of DeepSeek and uninstall the application if they have downloaded it. They also advise against using the DeepSeek applications (mobile app, desktop application, or web browser version) and will work to block access to the applications on the UBC network. UBC recommends using alternative AI models with better security practices instead. Learn more.
QwQ-32B: Alibaba’s Advanced Reasoning Model
Alibaba’s AI division, Qwen, introduced QwQ-32B, a powerful 32-billion parameter open-source reasoning model that matches the performance of larger models like DeepSeek’s R1. The new model specializes in resolving STEM inquiries using Reinforcement Learning (RL) to enhance its reasoning capabilities. Due to its comparatively small size, the model can be run on a user’s personal computer. Read more.
Try out QwQ-32B with Qwen Chat here.
Google’s Gemma 3: A Lightweight High-Performance AI
Google unveiled Gemma 3, an open-source AI model that can operate efficiently on a single GPU. The model is also capable of processing multimodal inputs and includes multilingual support. Gemma 3 reportedly outperforms competitors with larger models like Facebook’s Llama3-405B and OpenAI’s o3-mini while using a single GPU. Learn more.
Try out Gemma 3 with Google AI Studio here.
AI2 Launches OLMo2 32B Language Model
The Allen Institute for Artificial Intelligence released OLMo2 32B, a fully open model that outperforms GPT-3.5 Turbo and GPT-4o mini on a variety of academic benchmarks. OLMo2 32B is unique in that it delivers results on par with other open-weight models but with a significantly lower training cost. OLMo2 32B uses the same training recipe as the smaller scale models in the OLMo2 family. Learn more.
Try out OLMo2 32B here.
Mistral Small 3.1: Efficient Multimodal AI Model
Mistral AI released Mistral Small 3.1, a multimodal, multilingual model 24B parameter model. This model offers improved text performance and an expanded context window of up to 128k tokens, enhancing its ability to process and understand complex inputs. Its efficiency allows for high performance with reduced computational resources. Mistral Small 3.1 outperforms other models in its weight class such as Claude-3.5 Haiku and GPT-4o mini in performance benchmarks including text instruct, multimodal instruct, and multilingual processing. Learn more.
Cohere’s Command A: Enterprise-Focused Language Model
Cohere introduced Command A, the Canadian company’s most recent secure and high-performance language model tailored for enterprise applications, especially in STEM and coding inquiries. It matches or surpasses larger models like GPT-4o and DeepSeek-V3 in standard benchmarks while generating significantly less hardware costs. Command A is designed to assist businesses in tasks that demand security, accuracy, multilingualism. Due to its low compute requirements, users can run Command A on smaller devices. Read more.
Sesame’s Research on Conversational Voice
Sesame is conducting research to overcome the “uncanny valley” in conversational voice interactions with digital voice assistants. The team’s goal is to create digital companions with natural, expressive speech that conveys emotional intelligence, conversational dynamics, contextual awareness, and consistent personality. This involves developing models capable of understanding and adapting to context in real time. Researchers at Sesame report they have not yet overcome the “uncanny valley” in their digital voice assistants but are seeing improvements to their model’s friendliness and expressivity. Read more.
Try out a conversational voice demo here.
Manus: China’s Autonomous AI Agent
Butterfly Effect introduced Manus, a general AI agent designed to perform personal assistant tasks more efficiently than existing models. Manus uses multiple AI models such as Claude 3.5 Sonnet and Qwen in addition to independently operating agents to respond to inquiries. MIT Technology Review received an invite code for Manus and tested the model on a variety of assignments. The tests revealed that though Manus is user-friendly and performs well on analytical tasks requiring extensive research with limited scope. However, the model has a slower processing time than other models, crashes often, and sometimes struggles to process larger text inputs. Read more.
Learn more about Manu here.
Jenni AI: AI-Powered Writing Assistant
Jenni AI is an AI-powered writing assistant designed to help users read, write, and organize research efficiently. It offers features like real-time suggestions, citations, and customizable writing styles. Jenni AI assists in brainstorming ideas, structuring content, and refining language, making the writing process more streamlined. Jenni AI can even be used to produce a starting research document with user-provided inputs and can generate in-text references. Learn more.
Google Gemini: Personalized AI Responses
Google’s AI chatbot, Gemini, introduced new features that personalize responses based on users’ search histories as well as upgrades to older features. Building on the Gemini 2.0 Flash Thinking Experimental model, it can provide customized recommendations for restaurants, travel, and more by referencing past searches. Deep Research was upgraded using Gemini 2.0 Flash Thinking Experimental to create more detailed reports. Google also introduces more personalization to the Gemini app by connecting with other Google apps, services, and search history to generate individualized responses. Read more.
Russian Propaganda Influences Western AI Chatbots
A recent audit by NewsGuard uncovered that Russian disinformation campaigns have infiltrated Western AI chatbots, leading them to produce pro-Kremlin propaganda. The manipulation can be subtle, such as adding disinformation to the data sets AI models are trained on. This manipulation affects the accuracy and neutrality of AI-generated content, raising concerns about the susceptibility of AI systems to external influence. Users are urged to fact-check AI generated information either on their own or using third-party misinformation detectors. Read more.
Course Design Studio: A Hands-on Exploration of AI Course Design Tool – April 9, 2025
Nicole Ronan and Manual Dias from the Centre for Teaching, Learning and Technology (CTLT) at UBC are hosting an in-person studio teaching participants how to design courses using AI on April 9, 2025. The session will start with an overview of using AI for course design, step-by-step tutorials on how to use various tools, and time for individual questions. Register here.
Developing Critical Generative AI Literacy Through Challenge-Based Activities – April 16, 2025
On April 16, 2025, Lucas Wright and Manual Dias (CTLT) and Judy Chan (CTLT/LFS Faculty Liaison) are hosting an online event for participants, especially faculty, to learn about using GenAI in education and critically evaluate the GenAI outputs. The session includes interactive activities for participants to practice using GenAI tools responsibly in a variety of educational contexts. Register here.
Enhancing Teaching Materials with AI: Innovative Tools and Strategies – April 23, 2025
Lucas Wright from the CTLT is hosting an online interactive session on April 23, 2025 about using AI to create teaching materials. Participants will learn how to use several GenAI tools to create efficiently create lesson plans and other teaching materials. Register here.
Reimagining Assignments with Generative AI: Enhancing Creativity, Critical Thinking, and Authentic Assessment – April 30, 2025
On April 30, 2025, Lucas Wright from the CTLT will be hosting an online session teaching participants how to use GenAI to design and build assessments. Participants will have the opportunity to collaborate with each other and share insights on AI-enhanced assignments. Register here.
UBC Author Edit Pages
Tool of the Week: UBC Author Edit Pages
What is UBC Author Edit Pages?
UBC Author Edit Pages is an open-source GitHub project developed by Rich Tape at the UBC CTLT that enables instructors and staff to directly edit and author custom pages within WordPress. It’s designed to streamline the content update process by offering a lightweight, user-friendly editing interface. UBC Author Edit Pages is available as a WordPress plug-in tool.
How is it used?
Upon installation, the tool integrates with supported platforms, allowing users to make changes without needing to dive into complex backend systems or rely on developers. It’s especially helpful for instructional teams managing course or department pages. The GitHub repository provides clear setup instructions and configuration details.
What is it used for?
This tool is primarily used to simplify the process of editing static content pages across UBC-hosted platforms. Whether you’re maintaining a course website or updating departmental info, UBC Author Edit Pages gives non-technical users the flexibility to make changes quickly and independently.
Try UBC Author Edit Pages 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 […]