|
For the first time, IMSA is offering a year-long Artificial Intelligence Sequence as a computer-science elective offered to juniors and seniors. Kavithaa Suresh Kumar and CS colleagues have designed a course that centers on questions of how computers learn from data, recognize patterns, and make predictions. Students don’t just write code, but build, test, and interpret models in a curriculum that connects mathematical concepts from linear algebra, probability, and calculus to real-world datasets and modern Python libraries.
In AI-1, students get hands-on with authentic, messy data such as library collections, restaurant menus, and environmental data, using supervised and unsupervised methods to find structure in unlabeled data to move toward understanding of machine-learning workflows. They build on those foundations in AI-2 and apply convolutional neural networks for vision and patterns as well as natural language processing and LLMs.
Students design personalized AI projects associated with domains they care about—medicine, finance, sports, environmental science, the arts—and pursue a question that matters to them, both building data sets and building models to generate insights and productions. By understanding and applying the underpinnings of AI, they demystify the technology and prepare to help build it in the future.
Tool Spotlight: Deep Research for Unit Planning
In the last year, each of the major AI chat platforms has introduced a “Deep Research” mode that searches more comprehensively and arranges ideas more systematically than more basic search modes. These modes pursue computation for a longer period of time than an ordinary query or web search—up to ten minutes or more—and output a detailed report filled with citations and linked sources. Its hallucination rates are significantly lower than a typical LLM query.
The most obvious use of such a tool is to produce reports, but the tools can also help teachers identify promising resources and texts to incorporate into their curriculum. For example:
In the next three weeks, my class has days focused on speaker,
addressee, imagery, metaphor, sound, rhythm & meter, allusion, and fixed
forms. For each of those days, give me a list of 8 poems I should I
consider assigning—three of them canonical, three of them from
1950–2000, and 2 of them from 2000–the present.
I’m designing a unit on DNA and RNA with focuses on DNA structure and
base pairs, replication, trnscription, RNA types, translation, and gene
expression. For each of those topics, evaluate video resources that might
be helpful to advanced high-school juniors—list 5 potential videos for each
topic, then evaluate them for suitability for my class.
Each of the platforms’ research modes will be accessible from their tool menus. In ChatGPT, the button looks like this:
| 
