Роботи, що індексуються в Google Scholar
SINGLE-CHANNEL ADVISORY SYSTEMS AS INFORMATION SINKS: A DATA PROCESSING INEQUALITY BOUND ON EXECUTIVE DECISION QUALITY
Igor Ivitskiy
Опубліковано 24.04.2026
Як цитувати
Ivitskiy, I. (2026). SINGLE-CHANNEL ADVISORY SYSTEMS AS INFORMATION SINKS: A DATA PROCESSING INEQUALITY BOUND ON EXECUTIVE DECISION QUALITY. Матеріали конференцій МЦНД, (24.04.2026; Біла Церква, Україна), 66–68. https://doi.org/10.62731/mcnd-24.04.2026.001
Завантаження
Дані завантаження ще не доступні.
Авторське право (c) 2026 Igor Ivitskiy
Ця робота ліцензується відповідно до Creative Commons Attribution 4.0 International License.
Анотація
The contemporary executive who consults a single large language model as a strategic advisor is not, as marketing language often suggests, augmenting cognition with an objective counterweight. That person is constructing a sequential information channel with no independent error-correction mechanism. This abstract argues, on strictly information-theoretic grounds, that such single-channel advisory systems operate as information sinks: architectures in which the mutual information between the true state of the environment and the final decision is bounded above by a monotonically non-increasing sequence of processing stages. The bound is not a heuristic. It is a direct consequence of the data processing inequality.Посилання
- Cover T. M., Thomas J. A. Elements of Information Theory. 2nd ed. Hoboken: Wiley, 2006. 776 p.
- Sharma M., Tong M., Korbak T. et al. Towards understanding sycophancy in language models. Proceedings of the International Conference on Learning Representations (ICLR 2024). 2024. 32 p.
- Pronin E., Lin D. Y., Ross L. The bias blind spot: perceptions of bias in self versus others. Personality and Social Psychology Bulletin. 2002. Vol. 28, No. 3. P. 369–381.
- Page S. E. The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies. Princeton: Princeton University Press, 2007. 456 p.
- Du Y., Li S., Torralba A., Tenenbaum J. B., Mordatch I. Improving factuality and reasoning in language models through multiagent debate. Proceedings of the 41st International Conference on Machine Learning (ICML 2024). 2024. 17 p.
- Ivitskiy I. The M.A.T.H. framework: a first-principles approach to quant marketing in high-uncertainty environments. Zenodo. 2026. URL: https://doi.org/10.5281/zenodo.18552246 (access date: 09.04.2026).
- Ivitskiy I., Savchenko D., Sydorenko D. Data integrity as the terminal constraint in AI-driven advertising: an information-theoretic analysis of conversion fraud and agentic threat evolution. Zenodo. 2026. URL: https://doi.org/10.5281/zenodo.18675362 (access date: 09.04.2026).
- Ivitskiy I. From playbooks to code: translating bidding strategy into rigorous objective functions and constraints. Global Trends in Science and Education: proceedings of the IX International Scientific and Practical Conference (Kyiv, October 5, 2025). Kyiv: Research Europe, 2025. P. 30– 33. URL: https://researcheurope.org/wp-content/uploads/2025/10/re-05.10.2025-30-33.pdf (access date: 09.04.2026).