Academic Publications

Emerson Johnston and Sunny X. Liu, “Topic-Sensitive Differentiation and Algorithmic Equivalence on Short-Form Video Platforms: A Systematic Review of Visual Frameworks.” Computers in Human Behavior Reports, July 2026. Read here

This systematic review examines how content recommendation algorithms on short-form video platforms handle different types of content. Reviewing 148 algorithmic models from 123 studies published between 2015 and 2025, we distinguish algorithmic equivalence — applying one recommendation strategy across all content types — from topic-sensitive differentiation, where systems adjust strategy by content theme. We find that 142 of 148 models (95.9%) exhibit algorithmic equivalence, and argue that this one-size-fits-all approach may obscure public-interest content such as political material. The coded dataset is openly available at https://doi.org/10.17605/OSF.IO/349K7.

Emerson Johnston, “Anticipatory identity control: Upstream governance on digital platforms” Dialogues of Digital Society, vol. 1, no. 3, September 2025. Read here

This commentary examines the entrenchment of authoritarian governance through the algorithmic architectures of major digital platforms. As identity formation increasingly occurs in online spaces, the capacity to explore, articulate, and inhabit non-normative or politically inconvenient subjectivities is contingent upon the technical affordances and moderation protocols of the platforms that mediate digital public life. Authoritarian regimes have capitalized on this dependency, embedding ideological control both in content policy and in the infrastructural substrate of platforms themselves. This commentary argues that such infrastructural interventions constitute a form of anticipatory governance: preemptively structuring the field of possible selves to align with state-sanctioned norms. In doing so, they reshape digital identity not as an open-ended process, but as a site of containment, coercion, and control.

Emerson Johnston, “When Regulation Backfires: Cultural Infrastructure and the Mass Migration from TikTok to RedNote.” Proceedings of the 12th International Conference on Communities & Technologies (C&T ‘25), July 2025, pp. 28–43. Read here

This paper examines how the 2025 U.S. ban on TikTok triggered an unexpected policy backfire when millions of American users migrated to RedNote (Xiaohongshu), another Chinese-owned platform. Following the Supreme Court’s decision in TikTok Inc. v. Garland, over 4 million U.S. users — primarily digital opinion leaders — deliberately chose RedNote as an act of political protest, creating unprecedented cross-cultural connections that amplified rather than reduced Chinese digital influence.

Policy Publications

Emerson Johnston and Amy Zegart, “Update: DeepSeek AI and the Great Talent Competition.” Technology Policy Accelerator, Hoover Institution and Stanford Institute for Human-Centered Artificial Intelligence (HAI), June 2026. Read here

This update extends our April 2025 analysis from DeepSeek’s first five foundational papers to all seven, adding V3.2 (December 2025) and V4 (April 2026). The author pool has grown from 223 to 356 researchers. The core findings from the original analysis hold and, in most cases, sharpen: DeepSeek’s talent remains largely developed within China, with U.S.-trained researchers making up roughly the same share of the expanded cohort as before.

Emerson Johnston and Amy Zegart, “Policy Implications of the DeepSeek AI Talent Base.” Stanford Institute for Human-Centered Artificial Intelligence (HAI), May 2025. Read here

A companion piece to the Hoover Institution white paper, this article distills the core policy takeaways for U.S. decision-makers. It emphasizes the strategic relevance of talent development pipelines, the shifting landscape of global AI expertise, and the importance of long-term investment in domestic research ecosystems.

Emerson Johnston and Amy Zegart, “A Deep Peek into DeepSeek AI’s Talent and Implications for US Innovation.” Technology Policy Accelerator, Hoover Institution, April 2025. Read here

This white paper investigates the emergence of DeepSeek AI, a Chinese research firm whose rapid advancements in large language models have challenged dominant assumptions about American leadership in frontier AI. In it, we analyze the institutional trajectories and academic records of the over 200 researchers involved in DeepSeek’s foundational work. Our findings reveal a self-sustaining pipeline of highly skilled researchers largely developed within China, with limited long-term reliance on Western institutions. With this in mind, we argue that U.S. policymaking must address not just compute and infrastructure disparities, but the broader erosion of America’s global talent advantage in artificial intelligence.

Other Writing

Emerson Johnston, “Nothing of Me is Original…” Scholar Insight, Knight-Hennessy Scholars, August 2025 Read here

This short piece documents my journey as an adoptee who studies identity.

In the Press

  • DigiTimes, “US Chases AI Talent Abroad as China Strengthens Its Homegrown Pipeline,” July 13, 2026. Read article

  • Wall Street Journal, “How China Is Girding for an AI Battle With the U.S.,” July 28, 2025. Read article

  • Bloomberg Opinion, “Can China Compete in the AI Talent War?” July 9, 2025. Read article

  • The Decoder, “US Think Tank Warns of Reverse Brain Drain in China’s AI Sector,” May 6, 2025. Read article

  • Semafor “Reports: US losing edge in AI talent pool,” May 2, 2025. Read article

  • Chosun Biz, “54% of DeepSeek AI Researchers Trained Only in China Raises Concern for US Technology,” May 2, 2025. Read article

  • TipRanks, “U.S. Tech Risks Losing AI ‘Brains Race’ With China as American Dream Fades,” April 30, 2025. Read article

  • South China Morning Post, “Could Chinese AI Scientists Threaten US Tech Dominance? Study of DeepSeek Team Gives Clues,” April 29, 2025. Read article

  • Washington Times, “Investigation: China’s DeepSeek Finds America Losing Edge in Tech Talent,” April 23, 2025. Read article