What You Should Know

AI In Education

• Researchers from MIT conducted a study that “explores the neural and behavioral consequences of LLM-assisted essay writing”. Divided into three groups (LLM-users, search engine users, and brain-only users), the LLM-users demonstrated weaker brain connectivity. 83% having difficulty quoting sentences from their essays, but still maintained a sense of ownership of and satisfaction with their essays.

AI In Research

• A preprint from Stanford finds that clinicians using AI as a collaborator on diagnostic tests are more accurate than when using traditional resources. Clinicians achieved accuracies of 85% when the AI provided an initial opinion and 82% when the AI opinion followed the clinician’s assessment, compared to 75% when using traditional resources.
• Mathematicians tested reasoning model o4-mini with some of the world’s hardest solvable problems, to impressive results. One researcher reported “This is what a very, very good graduate student would be doing”, while also warning answers provided by o4-mini are so confident that they warrant additional scrutiny.

AI Current Events

• Copyright challenges continue to rise for AI companies, with a variety of media companies, including the likes of Disney and Universal, suing the AI Image company Midjourney for copyright infringement. Copyright issues continue to be a major concern for AI companies as some models have been found to reproduce exact passages from copyrighted works like The Hobbit and Harry Potter. This verbatim copying undermines AI companies’ legal defense that their models only learn general patterns from training data, a practice that would likely qualify as fair use.
• An AI Innovation and Law Fellow at the University of Texas argues that judges and state legislators are currently ill-equipped to manage AI risks due to a lack of resources and knowledge. A response from an Assistant Professor at Touro Law Center asserted that “liability represents the most suitable policy for addressing risks posed by AI systems”.
• A Gallup poll reports AI Use has nearly doubled at work, increasing from 21% to 40% over the last two years, with adoption increase coming primarily from white-collar roles.
• McKinsey reports that while nearly 8 in 10 companies report using Gen AI, they also report no significant bottom-line impact resulting in a “gen AI paradox”.

Weekly AI Tip

1. Guidance – What role is the AI filling? What guidelines does it need to follow? What provided information/documentation/data does it need?
2. Audience – Who is the intended output for? Yourself? Your class? A group of professionals? What is their knowledge/skill level?
3. Purpose – Why do you need this? What is the purpose of what you are trying to generate?
4. Output – How do you want your output structured? Do you have a template? Does it need to fit a word count? Does it need to be at a certain reading comprehension level?

AI Spotlight

• Diagnostic Application of Machine Learning Models for Predicting Feline Infectious Peritonitis
• Identifying Dogs at Risk of Antimicrobial Resistant Infections with Machine-learning: Escherichia coli Bacteriuria in Dogs
• Machine Learning Tools to Identify Cats with Bacterial Cystitis When Exhibiting Lower Urinary Tract Signs

Upcoming AI Events

• The Building Custom GPTs and Google Gems Workshop focuses on creating personalized and functional AI tools using ChatGPT and Google Gemini. Participants will learn to develop AI tutors, AI generators for assessment content, and AI writing/editing assistants. The workshop is designed so each attendee leaves with their own functional custom AI and the foundational knowledge to build additional tailored solutions. RSVP