The ThinkAI Initiative

Purdue College of engineering (COE) aims to lead the AI education in the United States by integrating AI into the curriculum through our ThinkAI Initiative. The educational goal of ThinkAI is to ensure that Purdue COE graduates will have competence in the following aspects of AI:

  • ThinkAI-Foundation: Broadly refers to the following skills
    • Problem Formulation: Abstract complex physical world problems.
    • Computational Thinking: Think algorithmically and procedurally. Know how to code.
    • Probabilistic Thinking: Think probabilistically; Learn from data; Assess likelihood.
  • ThinkAI-Partnership: Broadly refers to the readiness to utilize large language models for
    • Learning: Use AI to learn new topics
    • Working: Use AI to complete tasks more effectively

Successful implementation of ThinkAI will promote COE graduates to meet the minimum requirement in AI competence, and to improve their competitiveness in the job market.

Why these two skills? Will they be outdated?

ThinkAI-Foundation and ThinkAI-Partnership focus on the fundamentals. There is no doubt that today’s AI tools will be outdated one day, perhaps much quicker than our imagination. For example, ChatGPT 5.2 (2025) is already orders of magnitude better than ChatGPT 2.0 (2023), and two decades ago Python was outnumbered by MATLAB but today every top school and every tech company is using Python.

The above two skills are specifically structured to be tool-agnostic:

  • ThinkAI-Foundation: The focus here is the basic knowledge in the mathematical and programming tools underpinning AI models. The emphasis here is on understanding the principles of these tools, and developing a knowledge base so that the students can
    • Formulate problems from the real world to interact with the AI;
    • Write code and debug; and think algorithmically. The programming language is unimportant, but it must be a real programming language.
    • Make informed engineering decisions-based data; Think probabilistically by estimating the likelihood of events
    • Integrate AI building blocks inside the design process.
  • ThinkAI-partnership: There are three aspects here:
    • How to use the computer as a personalized assistant to help students learn new concepts, formulate problems, and brainstorm ideas.
    • How to distill the correct information from hallucination through principled cross-validation.
    • How to acknowledge the limitations of AI assistants and be aware of the ethical responsibilities.

Hierarchy of AI Learning Experience

The ThinkAI Initiative will make AI go into the curriculum. A traditional 3-credit-hour course focusing specifically on mathematics and programming of deep neural networks would not serve the purpose. To this end, we will distribute the learning experience across the years and will be integrated into the existing curriculum.

The new AI curriculum will consist of four levels of learning experience:

  1. Exposition (Level 1). The goal of Exposition is to get students exposed to the three required thinking skills. The target audience are the freshmen, through the existing FYE program. Students completing this level will have basic exposures to AI. They will also receive basic training in computer programming, statistical data analysis, and experience working with AI assistants. The emphasis at this level is exposure.
  2. Knowledge (Level 2). The goal of Knowledge is to provide students with the essential technical background so that they know the basic principles of how computing, statistical analysis, and AI-assistant work. Students completing this level should be able to briefly explain the principles behind these tools.
  3. Subject (Level 3). The goal of Subject is to provide students with the opportunity to experience AI in their own disciplines. For example, Civil Engineering students can Python to write codes to visualize traffic control patterns. Another example, Material Engineering can ask students to present a case study of optical lithography with the help of AI-assistants. Students completing this level will be able to demonstrate the usage of AI tools in their studies.
  4. Hands-On (Level 4). The goal of Project is to allow students to apply their AI skills in real-world projects. Given the depth and breadth of our undergraduate project infrastructures in COE, students participating in any research projects, senior designs, corporate-sponsored projects, challenges, VIP, and EPICS, can demonstrate how AI is used to improve their designs and workflow.