4. Perceived Changes in
Skills Needed for Business Students
More Needed Skills
• Integrity
• Strategic vision
• Ability to inspire
• Negotiation
• Interpersonal skills
• Oral communication
Less Needed Skills
• Specific language skills
• Written communication
• Quantitative analysis
6. Productivity Effects of AI in Professional Writing
Control Group:
No ChatGPT
Treatment
Group:
ChatGPT
Quality: 18% higher
Time spent: 40% lower
Job satisfaction: higher
Lower ability workers
benefit most
Source: Noy and Zhang (2023), Science, DOI: 10.1126/science.adh2586
8. AI and Creativity/Innovation
200 randomly
selected
product pitch
ideas from
MBA students
200 product
pitch ideas
generated by
GPT-4
vs.
9. AI and Creativity/Innovation
MBA
Students
GPT-4
Source: Terwiesch, C., & Ulrich, K. https://www.wsj.com/tech/ai/mba-students-vs-chatgpt-innovation-679edf3b
Time Spent
Days
An hour
Quality of Ideas
Purchase intention 40%
Purchase intention 49%
Of the Top 40 Ideas
5 from MBAs
35 from ChatGPT
14. Business Ethics and Ethical Use of AI
• Properly crediting AI use
• Privacy concerns
• Individuals
• Proprietary data
• Plagiarism/Intellectual property issues
15. Other Curricular Opportunities
• Personalized learning guidance and support
• Experiential learning courses
• Students help organizations assess how they can effectively use AI
• Courses on functional use of AI (e.g., in marketing, in finance)
• Executive education
18. Accounting Students vs. AI
Human
Students
GPT 3.5
Student
Performance Data
• 28,000 questions
• 186 institutions
• 14 countries
19. Accounting Students vs. AI
76.7%
47.5%
56.5%
0%
20%
40%
60%
80%
100%
Students GPT-3.5 GPT-3.5
(partial
credit)
• GPT performs better
• True/False questions
• Multiple choice questions
• Questions in English
• AIS and audit questions
• GPT performs worse
• Short-answer questions
• Workout questions
• Tax, financial, and managerial
accounting questions
Source: Wood et al. (2023) Issues in Accounting Education
20. How Well Does GPT Do on Standardized Tests?
Source: https://openai.com/research/gpt-4
21. Can We Reliably Detect
AI-generated Work at Scale?
“In short, no. While some (including OpenAI) have
released tools that purport to detect AI-generated
content, none of these have proven to reliably
distinguish between AI-generated and human-
generated content.”
Source: https://help.openai.com/en/articles/8313351-how-can-educators-respond-to-students-presenting-ai-
generated-content-as-their-own
— OpenAI
22. Changing Educational Assessment
• Increase expectations
• Exams
• Oral exams
• Administer in settings without access to technology (e.g., in class vs. take
home)
• Make less use of the types of questions GPT does well on
• Switch to or put more weight on non-exam assessments
• Presentations with Q&A to check understanding
• Class participation
• Shorter in-class quizzes without access to technology
24. AI and Academic Research
• Research on AI and business education
• Research on AI and business practice
• Use of AI in research more generally
• Synthesize existing literature
• Editing/formatting
• Regulation of AI
Editor's Notes
Dean of a business school
Graduating roughly 1800 students each year
We want them to be well equipped to add value to the organization that will hire
Note: ChatGPTs ability to do math is significantly enhanced by the addition of the Code Interpreter features available through ChatGPT+ or ChatGPT Enterprise. See link below for recent study.
https://arxiv.org/abs/2308.07921
Suggestion: If you have a ChatGPT+ account, do a quick demo with data analysis and some math with optimization. If interested I can set up an example.