Workshop of Digital Transformation Researchers
The second Joint Scientific Seminar for digital Transformation researchers was held with the support of the Academic Staff Reserve of the Higher School of Economics. Speakers included Mikhail Romashikhin from the Educational laboratory of computer-aided design systems (HSE University) and Maria Yudina from the International Laboratory for Digital Transformation in Public Administration systems (HSE University). Participants discussed applied and educational issues of digital transformation.
Mikhail Romashikhin made a presentation about the development of systems on a chip (SnC). It is a system consisting of several computing cores, accelerators, for example, machine learning units or a video processor, memory, interfaces, and much more. A promising architecture of modern multicore SNCs is a network on a chip (StnC). Mikhail Romashikhin showed how the logic of SNCs and STNCS were developed in hardware description languages, how they are hardware prototyped using specialized FPGA chips. He demonstrated the operation of test algorithms during prototyping.
Such solutions can be used for the development of domestic microelectronics, as well as for import substitution of unavailable CAD and ready-made solutions. The use of STNCS simplifies the development of SNCs, which will allow the use of foreign and domestic solutions in the development of modern domestic microcontrollers and multicore processors. Colleagues supported Mikhail Romashikhin's work, noting its importance for import substitution in the field of digital transformation.
Maria Yudina shared with her colleagues the project of scientific research seminars on the sociology of technology. A series of seminars has been developed for the Master's program "Population and Development" (HSE University). The program trains specialists in public and municipal administration. The aim of the seminars is to teach students how to analyze advanced technologies, taking into account social opportunities and risks, in order to improve the quality of decision-making. Students will try research in the digital transformation sphere and will defend projects in front of a group, evaluating each other and training presentation skills. The meeting participants supported the project, discussed the problems of students using AI, grade inflation and the difficulty of ensuring fair mutual assessment of students.
In particular, Angelina Yudina, a lecturer at the HSE Department of Mathematics and a research intern (HSE University), highlighted the key risks associated with the use of anonymous peer review. The main problem is the distortion of motivation, which leads to two negative scenarios:
1. Deliberate overestimation within the framework of "agreements" or because of personal sympathies.
2. Unjustified underestimation of grades. For example to artificially improve one's own position in the academic ranking.
The result of such practices is a shift in final grades and a decrease in the objectivity of the entire assessment system. Possible solution: to make the assessment non-anonymous for the teacher with random verification of the grade. If a student's assessment of a fellow student is very different from the teacher's position, then the student who makes such an assessment may lose points for his performance. This approach creates a system of incentives for objective assessment, increases the involvement of students in the projects of their classmates, and also allows them to develop critical thinking and the competence of reasoned assessment.
Young teachers noted the weakness of AI in technical specialties. Generative models often produce delusional solutions. If a student manages to get the right result with the help of AI, most often one cannot repeat it in the classroom or explain it, which teacher can easily detect.