The "New" Approach to Pedagogy in AI Age 

Originally posted: August 31, 2026

Key Points

• AI prompts all educators in the world to revise their approach to knowledge evaluation; Good performance on homework is no longer an indicator of learning.  

• The new approach needs to be able to extract the true signal of learning and be scalable to cover the growing demand in education. 

• With AI, the signaling cost of producing the learning artifact becomes negligible; a verifiable longitudinal trace of work carries more information about the student's capabilities. 

Introduction

Recently, I’ve been tuning in to discussions about revising pedagogical approaches in the age of AI. Among many academic discussions, I'm also interested in increasing the signaling value of education for employers.

The biggest effect of AI on the education industry is that it makes traditional learning artifacts cheaper to produce (e.g., homework, essays, labs, etc.). The signaling value of students' capabilities weakens. As it becomes harder to assess true knowledge, the informational value of traditional educational artifacts from the employers' perspective decreases. Hence, the educators' role in helping students to improve the quality of their individual signals in the economic marketplace becomes central.

That's why I have been rethinking the value of snapshot measures, such as one-time exam performance, versus the value of learning traces, such as repeated performance and development over time. A longitudinal, verifiable trace of work can provide a richer signal because it documents not only what a student produced, but how their capabilities developed over time. Arguably, the trace gives an employer more evidence about whether performance is repeatable rather than accidental. 

Below, I describe the practical pedagogical techniques that could help build this capability trace.

#1 Flipped classroom

The flipped-classroom approach is to come to class with a baseline level of understanding that makes conversation participation more accessible for the student. The student's job is to come prepared to contribute to the class. This is where AI in the form of LLMs may be of the greatest help. LLMs, as information-synthesis agents, reduce the cost of entry into the area of knowledge. In a sense, they equalize access to knowledge based on attention span. Equipped with this tool, a student should feel more comfortable preparing for the class independently. If explanations are available almost for free, coming to class with zero understanding becomes increasingly difficult to justify. 

The instructor's role here is to coordinate the individual preparation by providing a set of specially designed "fine-tuned" prompts to ensure less heterogeneity in the output among the different LLMs and geographically dispersed students. Afterward, the instructor's scarce classroom time can be reprioritized from explaining what something is toward discussing why it matters, when it fails, how competing explanations differ, and how it should be applied.

#2 Project work

Referring to my point in the introduction, one of the central components of the class is the elucidation of the student's capabilities. The student's job is to produce artifacts that can live beyond the classroom and signal their acquired capabilities to employers. Project work is one of the widely used techniques to put students' capabilities to the test, especially their ability to synthesize multiple skills. What becomes particularly important in the age of AI is the public and persistent nature of the artifact. Public visibility creates stronger incentives for effort, revision, and attention to quality because the work is evaluated beyond the boundaries of the classroom.

The instructor's role here is to provide avenues for the public visibility of the work and lead the students to its quality enhancement. A GitHub repository, a project website, a public analysis, a LinkedIn article, and a YouTube tutorial video are publicly visible artifacts that can be crafted and improved under the supervision of the instructor. This work later paves the way for the student's development and a verifiable track record of performance.

#3 Two-stage exams

Finally, the classroom assessment still remains actual. But instead of compartmentalized individual exam work verifiable only by the instructor, the two-stage exam adds a necessary transparency component by making students' mastery visible through group participation. The student's job is not only to score high but also to interact and collaborate with peers and articulate what they learned through the argument. Such interaction closely mimics the work environment and elicits the critical thinking skill that is the aspiration of the educational system.

The instructor's role here is to choose the method of the examination, e.g., a two-stage exam, that de-antagonizes the use of AI and creates peer-to-peer visibility of each other's approach to learning and the level of outcomes. The innate peer accountability is another source of motivation to prepare and perform autonomously. 

Conclusion

In the age when outsourcing academic work to AI becomes almost costless, the quality of the schoolwork artifact is a less reliable proxy for the level of an individual's capabilities. In the AI age, the scarce signal is shifting from a snapshot performance score to the progress they can demonstrate over time. The instructor’s role, therefore, is to create opportunities for students to leave behind credible evidence of accumulated capability. The student's responsibility is not merely to score high, but to produce evidence of an acquired skill set over time. 

References:

1. MIT Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. (2026, August 13). Report of MIT’s Ad Hoc Committee on AI use in teaching, learning, and research training. Massachusetts Institute of Technology. https://aiandeducation.mit.edu/report/ 

2. Spence, M. (1973). Job market signaling. The Quarterly Journal of Economics, 87(3), 355–374.

3. Svoronos, T. (2018, January 16). 15 minutes on two-stage exams. https://teddysvoronos.com/2018-01-16-15-minutes-on-two-stage-exams/ 

Please cite this article as: 

Petryk, M. (2026, August 31). The New Approach to Pedagogy in AI Age. MariiaPetryk.com. https://www.mariiapetryk.com/blog/post-34