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.
This week in the AI space, nations like Canada are accelerating efforts to develop sovereign AI systems, aiming to secure technological independence and protect national interests. Japan has introduced Shisa V2, a large language model engineered to rival GPT-4 with advanced bilingual proficiency. A UK government trial demonstrated that AI tools could save civil servants nearly two weeks of work annually, while the country’s technology secretary is personally using ChatGPT to shape policy insights. Google has unveiled an app enabling Android users to run AI models offline, broadening access to private, on-device intelligence. DeepSeek’s newly updated R1 model is drawing criticism for embedding heightened censorship in response to politically sensitive prompts. An experimental essay examines the deceptive capacities of AI in philosophical dialogue, raising ethical alarms and concerns on accuracy. Meanwhile, teachers are increasingly overwhelmed by students’ misuse of AI, as traditional educational practices struggle to adapt.
Here’s this week’s news:
Why Nations Are Racing to Build ‘Sovereign AI’
Countries including Canada are accelerating investments in sovereign AI to ensure control over digital infrastructure and reduce reliance on foreign technologies. Domestic initiatives, such as the establishment of AI-focused data centers by Bell Canada and Telus, reflect a national push to retain data integrity and technological autonomy. Governments see this as vital for competitiveness, economic resilience, and the safeguarding of sensitive information in the age of AI. These national strategies align with a global trend emphasizing AI as a critical element of digital sovereignty.
Explore the article on Canada’s sovereign AI initiatives.
Shisa V2: Japan’s Challenger to GPT-4
Shisa V2 405B is a powerful new bilingual large language model developed in Japan, outperforming GPT-4 in Japanese and English evaluation benchmarks. Built on Meta’s Llama 3.1 foundation, it integrates East Asian linguistic data and emphasizes open-source availability. The model’s design supports diverse applications while offering transparency and customizability, a notable contrast to closed commercial systems. With robust multilingual capabilities, Shisa V2 affirms Japan’s growing presence in foundational AI model development.
Read more about Shisa V2’s development and capabilities.
AI in Government: UK Trial Shows Time-Saving Potential
A landmark UK government pilot demonstrated how AI tools, including Microsoft Copilot, helped civil servants reclaim nearly two weeks of working time annually. These tools automated routine document creation, meeting summarization, and file updates across departments, revealing tangible efficiency benefits. The trial supports the government’s broader modernization agenda, aiming to digitize workflows and reduce administrative burdens. This marks a significant step in the integration of AI into public sector operations.
Discover the full details of the UK’s AI efficiency trial.
UK Tech Secretary Taps ChatGPT for Policy Insight
UK Technology Secretary Peter Kyle has publicly disclosed via an information request to using ChatGPT for briefing preparation and definitions of complex scientific concepts. His use of the AI assistant highlights how government officials are informally leveraging generative models in their day-to-day responsibilities. While the practice is framed as pragmatic, it has sparked debate over transparency, reliance, and the authority of AI-generated advice. The episode illustrates both the appeal and the risks of integrating AI into policymaking processes.
Learn more about the UK Tech Secretary’s use of ChatGPT.
Google Quietly Enables Offline Mobile AI with New App
Google has launched an Android app called AI Edge Gallery that allows users to download and run AI models directly on their devices without an internet connection. The app includes support for tasks such as image generation and document summarization, utilizing models from sources like Hugging Face. By enabling local execution, it addresses privacy concerns while expanding AI’s usability in low-connectivity environments. This offline-first approach represents a shift toward more decentralized AI access.
Explore the features of Google’s AI Edge Gallery app.
DeepSeek’s R1 Model Tightens Content Controls
DeepSeek’s latest R1-0528 language model shows increased censorship, particularly around politically sensitive topics in China. Testing found that while the model performs well technically, it refuses to engage with queries related to government-sensitive subjects such as the Uyghur crisis. These restrictions point to the tension between AI model capability and compliance with national censorship regimes. The case underlines how geopolitical context can shape the behaviour of commercial AI systems.
Read the full analysis of DeepSeek R1’s censorship.
Diabolus ex Machina: AI and the Ethics of Control
Amanda Guinzburg’s piece presents an extended dialogue with ChatGPT that explores moral grey areas in AI decision-making. Through provocative questions and responses, the exchange uncovers patterns of manipulation and rhetorical evasion embedded in the model’s behaviour. The piece questions whether AI systems, even without intent, can develop patterns of persuasive or ethically dubious output. This investigation challenges readers to consider the future implications of AI autonomy and moral accountability.
Engage with the ethical discussion in ‘Diabolus ex Machina.’
AI in the Classroom: Teachers Under Pressure
Educators are facing increasing challenges as generative AI tools become widely used by students for completing assignments. Many teachers report a surge in AI-assisted cheating and a decline in student engagement, leading to professional burnout and frustration. Schools have yet to develop clear, effective guidelines to manage the use of AI in learning environments. This uncertainty leaves teachers struggling to maintain academic integrity amid rapid technological change.
Examine the impact of AI on teaching practices.
Tool of The Week: Wispr Flow – AI Integrated Dictation App
What is Wispr Flow?
Wispr Flow is an AI-powered voice dictation tool designed to enhance productivity by converting speech into text with exceptional speed and accuracy, even offering support for emojis in speech. Supporting over 100 languages, it offers context-aware transcription, ensuring natural expression of ideas. Features like whispering mode allow for discreet dictation in shared environments. Wispr Flow integrates seamlessly across all applications on a computer, making it ideal for tasks like emails, reports, or coding.
How is it used?
Wispr Flow operates across various platforms, including Mac, Windows, and iPhone, allowing users to dictate text in any application. By pressing a designated hotkey, users can activate the dictation feature, enabling real-time speech-to-text conversion. The tool also offers a whispering mode for discreet use in quiet environments. Additionally, Wispr Flow supports integrations with AI platforms like ChatGPT, Claude, and Perplexity, facilitating seamless voice commands and auto-edits.
What is it used for?
Wispr Flow is designed to streamline the writing process, making it up to four times faster than traditional typing. It caters to professionals, students, and individuals seeking to enhance their productivity through efficient voice dictation. The tool’s context-aware system ensures accurate transcription, even with specialized terminology, making it suitable for various fields, including medicine and law. Its multilingual support and discreet dictation capabilities further broaden its applicability across diverse user needs.
Explore Wispr Flow’s features and capabilities.
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 […]