#012 ML Юрий Яровиков. Как окончить ШАД, руководить Школой глубокого обучения и остаться в России

Value of the Information & Strength of the Argument

The value lies in the insider perspective on elite Russian AI education and career progression. The guest provides candid observations about SHAD’s entry requirements (heavy on olympiad math), the lack of need for prior programming, and the risks of not engaging with the community in person. His argumentation is coherent, but predominantly anecdotal: he supports his points with personal experience rather than systematic evidence. He also offers pragmatic advice for students, such as valuing community ties and accepting academic breaks. The discussion of Sber’s shift from pure research to business-oriented ML is illustrative of corporate constraints, but lacks comparative analysis or broader industry data.

Scientific Rigor, Source Quality, Title Accuracy

The title matches content well, though ‘how to graduate’ oversimplifies the narrative. The episode references several institutions (MIPT, SHAD, DLS, Huawei, Sber) without offering external verification. The description contains official links (e.g., dlschool.org, mipt.ru) that support the existence and role of these organizations, but the claims about curricula, teaching quality, and corporate intentions are unverified. No discordant sources are discussed. The presentation is professional, but the scientific rigor is limited by the format’s subjective nature. Public comments are not provided for analysis.

The title accurately reflects the episode's content, which covers the guest's path through SHAD, leadership of the Deep Learning School, and his decision to remain in Russia. Slight emphasis on 'how to' suggests actionable advice, but the discussion is more of a narrative.

5/10

The video is a personal interview offering subjective career anecdotes and opinions without verifiable data or reproducible analysis. Claims about difficulty, curricula, and company practices are based on individual experience, not objective metrics. The credibility rests on the guest's expertise, but the lack of supporting evidence lowers reliability.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:00Introduction: host presents Yury Yarovikov and his affiliations.
• 02:00Yuri's background: olympiad mathematics in Kazan, early exposure to math.
• 05:00Choosing MIPT (FPMI) based on advice from teachers; evolution of the faculty.
• 10:00Entering Yandex School of Data Analysis; interviews and mathematical focus.
• 20:00Experience at SHAD: workload, remote attendance, and value of community.
• 30:00Work at Huawei: computer vision research, Face ID, and corporate secrecy.
• 40:00Transition to Sber: business-oriented ML tasks, predictive modeling for marketing.
• 50:00Move to Sber's research lab: focus on Artificial General Intelligence.

• Deep Learning School official site — Referenced as the school Yuri heads and promotes.
• VK community for DLS — Mentioned as a community platform for DLS students.
• MIPT FPMI department page — Yuri's university faculty, described as evolved from the former 'FIVT'.
• Podcast anchor.fm episode — The episode hosting page mentioned in description.
• Podcast anchor.fm channel — Main podcast channel for all episodes.
• Apple Podcasts listing — Podcast distribution link.
• Google Podcasts feed — Podcast distribution link.
• Yandex Music album — Podcast distribution link.
• Telegram channel 'To be an ML specialist' — Mentioned as related resource.
• VK podcast public — Podcast's VK community.

• Deep Learning School official site — Confirms the existence and mission of the school that Yuri leads.