TeIAS Computer Science Summer School was Held

According to the Public Relations and Website Office of Khatam University, the TeIAS Computer Science Summer School, titled “Modern Software Production,” took place on Monday, September 14. The event, held in the Hakim Azam Building Amphitheater, featured the participation of Master’s degree students in the field, several keynote lectures, and a discussion panel.

The one-day course was organized by Dr. Mohammad Ghaffari, a faculty member of the Tehran Institute for Advanced Studies (TeIAS). The program began with Dr. Ehsan Khamespanah, a faculty member of the Department of Electrical Engineering and Computer Science at the University of Tehran, who provided introductory remarks on the position and importance of the programming profession. He stated: “There are two opposing, and sometimes contradictory, perspectives in this field: first, the public’s view of programmers, where they are generally imagined as transcendental figures whose activities have little relation to the actual challenges people face; and second, the programmers’ view of the public, whom they often see as people unable to effectively use the software that they and their colleagues design. This contradiction has existed from the beginning, and the gap between these perspectives has only widened over time.”

In his lecture, titled “Effective Test Case Production Using Large Language Models,” Dr. Khamespanah addressed the unique characteristics of Large Language Models (LLMs), adding: “When dealing with highly complex systems, the probability of failure increases. In such situations, testing these systems before deployment is crucial, and ensuring they function correctly is a difficult task.”

Following this, Dr. Yadollah Yaghoubzadeh, a faculty member of the Tehran Institute for Advanced Studies (TeIAS), delivered a lecture titled “Creativity in Large Language Models.” While providing definitions of creativity and LLMs, he noted: “Creativity is not limited to the creation of works of art; we also encounter creativity in the realms of science and knowledge. Furthermore, creativity is not exclusive to human works; we observe a wide diversity of creativity in nature as well. Based on Boden’s definition, a creative thing must possess three characteristics: novelty, surprise, and value.”

He added: “Today, we have reached a point where Large Language Models function similarly to humans in many respects. The question then is: Can these language models be creative? The answer is yes; to claim otherwise would be a contradiction. Another question is: How closely does the output of these language models resemble human creativity? Certainly, the nature of these creativities is different, and we need to integrate humans into this machine creativity. Culture, the human environment, and experience can be influential in this regard.”

Subsequently, Dr. Razieh Bahjati, an Artificial Intelligence specialist who joined the session online from the United Kingdom, attempted to answer the question, “Is AI Transforming Open-Source Software?”, by presenting practical and up-to-date examples.

Dr. Ali Ebnenasir, a Professor of Computer Science at the University of Michigan Tech, delivered a lecture titled “AI-Based Software Development Despite Resource Constraints.” He sought to answer: “How can we design neural networks so that they function effectively on computers with limited speed and memory, which also face energy constraints?”

Following this, Dr. Mahboubeh Samadi, a faculty member at Shahid Beheshti University, spoke on the topic “AI for Software Behavioral Model Extraction.” Referring to how children easily learn to use mobile phones, computers, and tablets at a young age, she raised the question: “Can a computer automatically learn and execute the behavior of a system?” In this context, she discussed the applications of using AI in reverse engineering in the modern world.

Additionally, Dr. Behnam Bahrak, an Assistant Professor at the Department of Electrical Engineering and Computer Science at the University of Tehran, delivered a lecture titled “Cybersecurity in the Age of Artificial Intelligence.” Citing the highly publicized recent attack by OpenAI agents on Hugging Face, he said: “The most critical issue regarding this attack was that no human agent issued the command for it. This issue could be more severe and alarming when an agent has access to both a company’s data and mastery over its various operations. Until yesterday, the question was: Can an attacker enter our system? But today, the question is: Can an AI cause us harm without having any malicious intent? Unfortunately, the answer is yes.”

Next, Dr. Mohammad Ghaffari, a faculty member of the Tehran Institute for Advanced Studies (TeIAS), presented a discussion on “Quality Engineering for Agent-Based Software.” The session concluded with a discussion panel featuring the participating professors, focusing on the question: “Will Artificial Intelligence replace software engineers?”

In response to a question regarding the future of IT jobs, Dr. Bahrak stated: “The need for specialists in specialized fields will never disappear.”

Regarding the same topic, Dr. Ebnenasir spoke about the capabilities LLMs possess in solving problems related to skill acquisition, adding: “As tools become more advanced, we move toward more complex problems to solve them.”

Dr. Ghaffari also remarked on how the learning gap has significantly narrowed compared to before and noted the opportunities this has created for growth.

In response to a question about the changes that will occur with the entry and expansion of AI into academia, Dr. Ghaffari said: “When there is freedom to use LLMs within the university environment, we can expect to cultivate more responsible students. Although achieving this is a time-consuming process.”

Similarly, Dr. Khamespanah noted that the entry of AI into education has made performing mundane tasks easier; therefore, higher precision can be expected from students.

Dr. Yaghoobzadeh also stated in this regard: “We must understand what is happening with the entry of AI into universities. If we cannot recognize ourselves as active participants within the loop, we will gradually lose our position.”

Dr. Ebnenasir concluded by stating: “On one side of this issue (the changes brought by the expansion of AI in academia) are the students, but on the other side is the academic system itself. It is important to determine what the academic system itself must do in this regard.”