{"id":"george-zeng","title":"George Zeng","content":"**George Xian Zeng** is the Chief Product Officer (CPO) of [NEAR Protocol](https://iq.wiki/wiki/near-protocol) and the General Manager (GM) of NEAR AI [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn). Appointed to the role in mid-2025, he leads the NEAR Foundation's strategic initiatives in decentralized and user-owned artificial intelligence [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29) [\\[4\\]](#cite-id-JLzdnI9JjXkcbs64). Before joining NEAR, Zeng was the Chief Operating Officer (COO) of the decentralized exchange [dYdX](https://iq.wiki/wiki/dydx) and co-founded the social e-commerce company Moonship [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29). His career also includes product leadership roles at Facebook, where he was instrumental in developing and scaling core advertising technologies [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn).\n\n## Education\n\nGeorge Zeng holds multiple advanced degrees from Ivy League institutions. He earned a Master of Business Administration (MBA) from The Wharton School, with a focus on Technology and Healthcare. Prior to his business studies, he completed a Master of Science in Computer Science with a specialization in Artificial Intelligence at the University of Pennsylvania. He received his undergraduate degree, a Bachelor of Arts (AB) in Economics with a focus on math, finance, and Chinese language and culture, from Princeton University. [\\[2\\]](#cite-id-030OMEEtfbjbMTKn)\n\n## Career\n\nZeng's career spans entrepreneurship, product management at major technology firms, and executive leadership in the [Web3](https://iq.wiki/wiki/web3) industry. He has founded multiple companies and is an active angel investor and startup advisor [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn).\n\n### NEAR Foundation\n\nIn mid-2025, George Zeng was appointed to a dual leadership role at the [NEAR Foundation](https://iq.wiki/wiki/near-foundation) as Chief Product Officer of [NEAR Protocol](https://iq.wiki/wiki/near-protocol) and General Manager of the newly formed NEAR AI division [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29). While some reports initially announced his title as Chief Growth Officer (CGO), his official and public-facing roles are CPO and GM [\\[4\\]](#cite-id-JLzdnI9JjXkcbs64) [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY).\n\nAs head of NEAR AI, Zeng is responsible for the foundation's strategic push into decentralized artificial intelligence, with a vision centered on creating \"user-owned AI\" [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29). His mandate involves leading the development of privacy-preserving AI products and infrastructure on the [NEAR protocol](https://iq.wiki/wiki/near-protocol), aiming to shift the ownership and control of AI agents from centralized corporations to individual users [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29) [\\[5\\]](#cite-id-9EbMOgccIc0dv6hz).\n\nUnder his leadership, NEAR AI launched \"IronClaw\" in February 2026. IronClaw is an open-source, verifiable, and secure runtime environment designed for \"always-on\" AI agents. The project aims to solve the security, compliance, and data ownership challenges that hinder the deployment of autonomous AI systems in critical workflows. At the time of its launch, Zeng stated:\n\n> \"AI agents are rapidly becoming part of core workflows, and critical questions around security, compliance, and data control remain unresolved. We’re extending NEAR AI’s commitment to secure, privacy-first infrastructure with IronClaw, giving developers and enterprises the foundation they need to deploy agentic AI systems responsibly without sacrificing control or transparency.” [\\[5\\]](#cite-id-9EbMOgccIc0dv6hz)\n\nIronClaw operates within an encrypted Trusted Execution Environment (TEE) and uses a defense-in-depth security model to isolate third-party tools, control network access, manage credentials securely, and prevent data exploitation. The system is designed to not collect any telemetry or analytics data, ensuring that user data remains within their local environment. [\\[5\\]](#cite-id-9EbMOgccIc0dv6hz)​\n\n### dYdX\n\nPrior to his role at NEAR, Zeng served as the Chief Operating Officer (COO) of [dYdX](https://iq.wiki/wiki/dydx), a prominent [decentralized cryptocurrency exchange](https://iq.wiki/wiki/decentralized-exchange) [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[4\\]](#cite-id-JLzdnI9JjXkcbs64). Sources indicate his tenure as COO was between 2022 and late 2025 [\\[2\\]](#cite-id-030OMEEtfbjbMTKn) [\\[6\\]](#cite-id-UERriVNpmFXgWX4l). This was not his first involvement with the company; he had previously served as an advisor and helper to dYdX from 2018 to 2019, during its earlier stages [\\[2\\]](#cite-id-030OMEEtfbjbMTKn).\n\n### Entrepreneurial Ventures\n\nZeng has a significant background as a founder and chief executive of multiple technology startups.\n\n#### Moonship\n\nFrom 2018 to 2022, Zeng was the Co-founder and CEO of Moonship, a social e-commerce company [\\[2\\]](#cite-id-030OMEEtfbjbMTKn). Moonship provided social group buying tools for merchants on the Shopify platform. The company's mission was to help merchants acquire customers through organic social channels, thereby reducing their dependence on increasingly expensive paid advertising on platforms like Facebook. Speaking on the motivation for founding the company, Zeng noted his experience at Facebook's ad division:\n\n> \"I built out the programmatic ad tech stack at Facebook and saw merchants spend 40-60% of their revenues on Facebook ads. We kept thinking that’s an absurd amount of money – and that it would be game changing if we can help merchants acquire customers via organic channels.\" [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29)\n\nTo build Moonship, Zeng assembled a team of engineers with experience from major tech companies, including individuals who architected Facebook's search engine and former Google employees. [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29)​\n\n#### AirCare Labs\n\nFrom 2012 to 2014, Zeng co-founded AirCare Labs and served as its CEO and Product Lead. [\\[2\\]](#cite-id-030OMEEtfbjbMTKn)​\n\n#### Groupon (China)\n\nEarly in his entrepreneurial career, in 2011, Zeng was involved in the international expansion of Groupon. He was a co-founder of the company's venture into China and was responsible for its business in Southern China as the Head of Sales for the region. [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn)​\n\n### Facebook (Meta)\n\nFrom 2014 to 2018, Zeng was a Product Lead at Facebook (now Meta) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn). During this time, he played a key role in building and scaling several core products within Facebook's ad technology stack. His contributions included work on the Ads API, Dynamic Ads, and Dynamic Creative Optimization. His work was credited with growing the ad tech stack he managed from its inception to over $1 billion in revenue. [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29)​\n\n### Investing and Advisory Roles\n\nIn addition to his full-time roles, Zeng has been an active angel investor and startup advisor since 2014, identifying himself as a \"Startup investor/helper\" [\\[1\\]](#cite-id-RSm75wXjtb1uAFFY) [\\[2\\]](#cite-id-030OMEEtfbjbMTKn). His advisory experience includes work with [dYdX](https://iq.wiki/wiki/dydx) from 2018 to 2019 and a consultant role at [Anchorage Digital](https://iq.wiki/wiki/anchorage-digital) in 2018. He was also a member of South Park Commons, a collective of technologists and entrepreneurs, from 2024 to 2025. [\\[2\\]](#cite-id-030OMEEtfbjbMTKn)​\n\n## Public Engagement\n\nZeng is a public speaker at technology and [Web3](https://iq.wiki/wiki/web3) industry events, including [NEARCON](https://iq.wiki/wiki/nearcon), the official conference for the NEAR ecosystem. In his talks, he discusses the vision and development roadmap for NEAR AI, the future of user-owned [AI agents](https://iq.wiki/wiki/ai-agents), and a concept he refers to as \"Agentic Commerce.\" [\\[3\\]](#cite-id-OLw4rNN5jNUAKb29)​\n\n## Interviews\n\n### Leadership and Organizational Development at dYdX #01\n\nIn Conversations Ep. 1 – George Zeng, released on April 18, 2023, on the YouTube channel Conversations, George Zeng outlines his perspectives on leadership, company development, decision-making processes, and career progression. The discussion is based on his professional background in finance, consulting, large technology firms, startups, and his executive role at [dYdX](https://iq.wiki/wiki/dydx).\n\n[YOUTUBE@VID](https://youtube.com/watch?v=wRmISNYdN08)\n\nDuring the interview, Zeng describes organizational growth at [dYdX](https://iq.wiki/wiki/dydx) as closely linked to recruitment standards. He states that the company has prioritized hiring individuals assessed internally as high performers and has, at times, left roles unfilled for extended periods while continuing the search process. He connects this approach to a longer-term planning horizon, indicating that certain operational decisions are evaluated over multi-year periods rather than immediate short-term outcomes.\n\nThe conversation also addresses Zeng’s approach to decision-making. He explains that earlier in his career he relied primarily on analytical and data-driven reasoning. Over time, he incorporated intuitive judgment into executive processes. He characterizes intuition as an initial assessment that emerges prior to structured analysis and states that such assessments are subsequently examined through data and reasoning. He also references structured methods of identifying emotional states as part of clarifying internal reactions before final decisions are made.\n\nIn comparing professional environments, Zeng distinguishes startup settings from sectors such as finance. He describes startup and crypto ecosystems as environments where participants often share long-term expectations about technological development. He contrasts this with industries he considers more competitive in structure, noting differences in incentive dynamics and collaboration patterns.\n\nRegarding early-stage company formation, Zeng identifies product relevance as a central variable. He states that companies frequently encounter difficulties when building products for which demand is limited or insufficiently strong. He refers to the importance of founders understanding the problems they aim to address and cautions against selecting projects solely on the basis of ease of implementation. He also discusses the concept that more complex operational challenges, including regulatory or infrastructure-related constraints, may require sustained effort to execute effectively.\n\nZeng comments on personal motivations influencing career decisions. He attributes part of his professional direction to family background and financial considerations. He identifies long-term financial stability and continued participation in company-building activities as recurring objectives. He further observes that individuals vary in their response to uncertainty, noting that some interpret uncertain conditions as risk, while others interpret them as opportunity.\n\nIn discussing executive leadership, Zeng emphasizes experiential knowledge developed through direct involvement in hiring, organizational structuring, and strategic planning. He indicates that this form of knowledge is cumulative and context-dependent. He also references interpersonal capabilities, including communication, trust development, and emotional regulation, as components of executive responsibility.\n\nAt the conclusion of the interview, Zeng states that increasing his technical understanding of [dYdX’s](https://iq.wiki/wiki/dydx) systems and infrastructure remains an ongoing objective. He notes that balancing technical engagement with executive oversight presents operational constraints, but considers technical familiarity relevant to long-term organizational planning.\n\nThe interview documents Zeng’s description of leadership as a combination of long-range planning, recruitment standards, structured decision-making processes, and alignment between individual disposition and professional responsibilities. [\\[7\\]](#cite-id-E6Rv40XAuzWPDvUy) \n\n### Decentralized AI and Privacy Infrastructure #02\n\nOn December 17, 2025, the YouTube channel SAIL Media published an interview recorded during NeurIPS 2025 featuring George Zeng, Chief Product Officer of [NEAR Protocol](https://iq.wiki/wiki/near-protocol) and General Manager of Near AI. The discussion focused on the development of decentralized artificial intelligence systems and privacy-preserving infrastructure.\n\n[YOUTUBE@VID](https://youtube.com/watch?v=ytA730eYb-8)\n\nDuring the interview, Zeng described Near AI’s work on decentralized and user-controlled AI systems. He stated that the organization is developing confidential inference products designed to process AI queries within Trusted Execution Environments (TEEs). These environments use hardware-based isolation mechanisms intended to restrict access to data during model inference. According to his explanation, this architecture is structured so that query data remains inaccessible to service operators and third parties while computation is performed.\n\nZeng referenced the historical corporate principle “Don’t Be Evil,” commonly associated with Google, and contrasted it with a model in which system design limits the ability of operators to access user data at a technical level. He indicated that centralized AI laboratories, including OpenAI and Anthropic, play an active role in advancing frontier models, while decentralized systems represent an alternative infrastructure model focused on privacy and distributed control.\n\nThe interview also addressed developer participation. Zeng noted that Near AI is offering confidential inference APIs intended for integration into third-party applications. He stated that current and planned development efforts include expanding support for model fine-tuning, training workflows, and mechanisms enabling users to transfer AI-related data across applications within secure execution environments.\n\nData portability was presented as a design objective. Zeng described an approach in which user interaction history and preference data could be retained under user control and made available across different software services, subject to privacy-preserving constraints implemented at the infrastructure level.\n\nIn discussing the broader AI landscape, Zeng commented on the relationship between open-source and closed-source frontier models. He observed that performance comparisons vary over time as new models are released. He indicated that long-term development may involve continued interaction between open and closed ecosystems, with knowledge and techniques influencing both.\n\nZeng also referred to [Illia Polosukhin](https://iq.wiki/wiki/illia-polosukhin), co-founder of [Near](https://iq.wiki/wiki/near-protocol) and co-author of the research paper Attention Is All You Need, noting his academic background and role within the organization.\n\nThe interview concluded with brief references to forthcoming announcements related to private chat tools and expanded confidential inference services for consumer, enterprise, and developer use cases.\n\nThe discussion reflects Zeng’s position that hardware-enforced confidentiality and decentralized system architecture represent one approach to addressing data control and privacy considerations in AI deployment. [\\[8\\]](#cite-id-kI3JyMbjJXYUAQZp) \n\n## Decentralized AI and Privacy Infrastructure\n\nOn December 17, 2025, the YouTube channel SAIL Media published an interview recorded during NeurIPS 2025 featuring George Zeng, Chief Product Officer of NEAR Protocol and General Manager of Near AI. The discussion focused on the development of decentralized artificial intelligence systems and privacy-preserving infrastructure.\n\nDuring the interview, Zeng described Near AI’s work on decentralized and user-controlled AI systems. He stated that the organization is developing confidential inference products designed to process AI queries within Trusted Execution Environments (TEEs). These environments use hardware-based isolation mechanisms intended to restrict access to data during model inference. According to his explanation, this architecture is structured so that query data remains inaccessible to service operators and third parties while computation is performed.\n\nZeng referenced the historical corporate principle “Don’t Be Evil,” commonly associated with Google, and contrasted it with a model in which system design limits the ability of operators to access user data at a technical level. He indicated that centralized AI laboratories, including OpenAI and Anthropic, play an active role in advancing frontier models, while decentralized systems represent an alternative infrastructure model focused on privacy and distributed control.\n\nThe interview also addressed developer participation. Zeng noted that Near AI is offering confidential inference APIs intended for integration into third-party applications. He stated that current and planned development efforts include expanding support for model fine-tuning, training workflows, and mechanisms enabling users to transfer AI-related data across applications within secure execution environments.\n\nData portability was presented as a design objective. Zeng described an approach in which user interaction history and preference data could be retained under user control and made available across different software services, subject to privacy-preserving constraints implemented at the infrastructure level.\n\nIn discussing the broader AI landscape, Zeng commented on the relationship between open-source and closed-source frontier models. He observed that performance comparisons vary over time as new models are released. He indicated that long-term development may involve continued interaction between open and closed ecosystems, with knowledge and techniques influencing both.\n\nZeng also referred to Ilia Polosukhin, co-founder of Near and co-author of the research paper *Attention Is All You Need*, noting his academic background and role within the organization.\n\nThe interview concluded with brief references to forthcoming announcements related to private chat tools and expanded confidential inference services for consumer, enterprise, and developer use cases.\n\nThe discussion reflects Zeng’s position that hardware-enforced confidentiality and decentralized system architecture represent one approach to addressing data control and privacy considerations in AI deployment.","summary":"George Zeng is a tech executive, Chief Product Officer at NEAR, and GM of NEAR AI. He previously served as COO of dYdX and held leadership roles at Meta and Red...","images":[{"id":"QmbbKa1tcmqUbFXqEFYw37BpSuKMyWtAx1BfpW5wjZ86DQ","type":"image/jpeg, image/png"}],"categories":[{"id":"people","title":"People in crypto"}],"tags":[{"id":"Founders"},{"id":"Developers"},{"id":"PeopleInDeFi"}],"media":[{"id":"QmbVatB2eAW9sijvMiRq7JdE2aL5Ue72amR7Thfgi8btoY","size":null,"name":"NC_SPEAKER_41-George-Zeng_NEARCON-1080x1080-1.jpg","type":null,"source":"IPFS_IMG"},{"id":"QmShjc3CzWve5RcMXTC6HiaV5kNGY3WYhdXrDQcEW5tANq","size":null,"name":"8pZQfbI3MG5F8KJq.jpg","type":null,"source":"IPFS_IMG"},{"id":"QmcD4qZCXst2B3rR8StUnDd1ncxHgZGXcztroRhXAjcBwK","size":null,"name":"Founder_George-Zeng.png","type":null,"source":"IPFS_IMG"},{"id":"https://www.youtube.com/watch?v=wRmISNYdN08","name":"wRmISNYdN08","caption":"","thumbnail":"https://www.youtube.com/watch?v=wRmISNYdN08","source":"YOUTUBE"},{"id":"https://www.youtube.com/watch?v=ytA730eYb-8","name":"ytA730eYb-8","caption":"","thumbnail":"https://www.youtube.com/watch?v=ytA730eYb-8","source":"YOUTUBE"}],"metadata":[{"id":"references","value":"[{\"id\":\"RSm75wXjtb1uAFFY\",\"url\":\"https://x.com/georgexzeng\",\"description\":\"George Zeng's X profile bio\",\"timestamp\":1772018276901},{\"id\":\"030OMEEtfbjbMTKn\",\"url\":\"https://rocketreach.co/george-zeng-email\\\\_13844546\",\"description\":\"RocketReach profile for George Zeng\",\"timestamp\":1772018276901},{\"id\":\"OLw4rNN5jNUAKb29\",\"url\":\"https://goldhouse.org/people/george-zeng/\",\"description\":\"Gold House A100 profile for George Zeng\",\"timestamp\":1772018276901},{\"id\":\"JLzdnI9JjXkcbs64\",\"url\":\"https://www.theblock.co/post/365214/30-key-crypto-hires-moves-and-exits-july-2025\",\"description\":\"The Block article on July 2025 crypto hires\",\"timestamp\":1772018276901},{\"id\":\"9EbMOgccIc0dv6hz\",\"url\":\"https://www.crypto-reporter.com/press-releases/near-ai-launches-ironclaw-a-secure-runtime-for-always-on-ai-agents-122807/\",\"description\":\"Crypto Reporter press release on IronClaw launch\",\"timestamp\":1772018276901},{\"id\":\"UERriVNpmFXgWX4l\",\"url\":\"https://www.mexc.co/en-NG/news/134801\",\"description\":\"MEXC News article on NEAR appointments\",\"timestamp\":1772018276901},{\"id\":\"E6Rv40XAuzWPDvUy\",\"description\":\"Conversations Ep. 1 - George Zeng\\n\",\"timestamp\":1772019487765,\"url\":\"https://www.youtube.com/watch?v=wRmISNYdN08\"},{\"id\":\"kI3JyMbjJXYUAQZp\",\"description\":\"The Future of Decentralized AI: George Zeng (Near AI) at NeurIPS 2025\\n\",\"timestamp\":1772019692039,\"url\":\"https://www.youtube.com/watch?v=ytA730eYb-8\"}]"},{"id":"twitter_profile","value":"https://x.com/georgexzeng"},{"id":"linkedin_profile","value":"https://www.linkedin.com/in/georgexzeng"},{"id":"previous_cid","value":"\"https://ipfs.everipedia.org/ipfs/QmNR4yFq1AVY7fLiiZqeyUCsP5RXES4pQ26MRVRXzXd6do\""},{"id":"commit-message","value":"\"Updated content and title for George Xian Zeng\""},{"id":"previous_cid","value":"QmNR4yFq1AVY7fLiiZqeyUCsP5RXES4pQ26MRVRXzXd6do"}],"events":[{"date":"2018-05-01","title":"Graduated MBA from Harvard Business School","type":"DEFAULT","description":"Completed his Master of Business Administration from Harvard Business School.","link":"https://www.mexc.co/en-NG/news/134801","multiDateStart":null,"multiDateEnd":null,"continent":null,"country":null,"id":"78598a70-3ecd-4c7f-b07c-270b942f6832"},{"date":"2018-01-01","title":"Co-founded E-commerce Company Moonship","type":"DEFAULT","description":"Co-founded and served as CEO of Moonship, a social e-commerce company providing tools for the Shopify ecosystem.","link":"https://goldhouse.org/people/george-zeng/","multiDateStart":null,"multiDateEnd":null,"continent":null,"country":null,"id":"dac543e0-c5ce-42cd-8519-6404f5691840"},{"date":"2024-01-01","title":"Became COO of dYdX","type":"DEFAULT","description":"Served as the Chief Operating Officer for the decentralized derivatives exchange dYdX.","link":"https://www.mexc.co/en-NG/news/134801","multiDateStart":null,"multiDateEnd":null,"continent":null,"country":null,"id":"347579d7-4601-4f37-a8a9-bad25fe42aac"},{"date":"2025-07-01","title":"Appointed CPO at NEAR & GM of NEAR AI","type":"DEFAULT","description":"Joined the NEAR Foundation as Chief Product Officer and General Manager of the NEAR AI division to lead its push into user-owned AI.","link":"https://near.foundation/blog/near-foundation-strengthens-leadership-and-enters-next-phase-of-growth-to-advance-ai-innovation/","multiDateStart":null,"multiDateEnd":null,"continent":null,"country":null,"id":"0357f3de-ff14-41d1-97e3-67e14fb4a9b8"}],"user":{"id":"0x8af7a19a26d8fbc48defb35aefb15ec8c407f889"},"author":{"id":"0x8AF7a19a26d8FBC48dEfB35AEfb15Ec8c407f889"},"operator":{"id":"0xd5893989b9952c6568a99F854795AcC5Ae480D56"},"language":"en","version":1,"linkedWikis":{"blockchains":[],"founders":[],"speakers":[]},"recentActivity":"{\"items\":[{\"id\":\"b717f81c-2e86-442f-94c4-92000aafd7ce\",\"title\":\"George Zeng\",\"description\":\"George Zeng is a tech executive, Chief Product Officer at NEAR, and GM of NEAR AI. He previously served as COO of dYdX and held leadership roles at Meta and Red...\",\"timestamp\":\"2026-02-25T11:45:40.523Z\",\"category\":\"People in crypto\",\"status\":{\"icon\":\"RiGlobalLine\",\"label\":\"Wiki Updated\",\"iconClassName\":\"text-green-500\"},\"user\":{\"name\":\"0x8AF7a19a26d8FBC48dEfB35AEfb15Ec8c407f889\",\"address\":\"0xd5893989b9952c6568a99F854795AcC5Ae480D56\"},\"button\":{\"label\":\"View Summary\",\"icon\":\"RiFileTextLine\"},\"summarySections\":[{\"title\":\"Title\",\"subtitle\":\"The title of the wiki was updated.\",\"variant\":\"modified\",\"changeCount\":1,\"changes\":[\"Changed title from 'people' to 'People in crypto'\"]},{\"title\":\"Summary\",\"subtitle\":\"The summary text was updated.\",\"variant\":\"modified\",\"changeCount\":1,\"changes\":[\"The summary text was truncated.\"]},{\"title\":\"Interviews\",\"subtitle\":\"A new 'Interviews' section was added with summaries of two interviews.\",\"variant\":\"added\",\"changeCount\":2,\"changes\":[\"Added subsection 'Leadership and Organizational Development at dYdX #01', detailing George Zeng's perspectives on leadership, company development, and decision-making processes. [[7]](#cite-id-E6Rv40XAuzWPDvUy)\",\"Added subsection 'Decentralized AI and Privacy Infrastructure #02', focusing on Near AI's development of decentralized and user-controlled AI systems. [[8]](#cite-id-kI3JyMbjJXYUAQZp)\"]},{\"title\":\"Decentralized AI and Privacy Infrastructure\",\"subtitle\":\"A duplicate section was added.\",\"variant\":\"added\",\"changeCount\":1,\"changes\":[\"Added a duplicate section titled 'Decentralized AI and Privacy Infrastructure' which repeated content from the 'Interviews' section.\"]},{\"title\":\"Events\",\"subtitle\":\"The date format for multiple events was updated for more precision.\",\"variant\":\"modified\",\"changeCount\":4,\"changes\":[\"Updated event date from '2018-05' to '2018-05-01'.\",\"Updated event date from '2018-01' to '2018-01-01'.\",\"Updated event date from '2024-01' to '2024-01-01'.\",\"Updated event date from '2025-07' to '2025-07-01'.\"]},{\"title\":\"Videos\",\"subtitle\":\"Two new videos were added to the wiki.\",\"variant\":\"added\",\"changeCount\":2,\"changes\":[\"Added YouTube video: https://www.youtube.com/watch?v=wRmISNYdN08\",\"Added YouTube video: https://www.youtube.com/watch?v=ytA730eYb-8\"]},{\"title\":\"References\",\"subtitle\":\"New references were added.\",\"variant\":\"modified\",\"changeCount\":2,\"changes\":[\"Added reference: 'Conversations Ep. 1 - George Zeng'\",\"Added reference: 'The Future of Decentralized AI: George Zeng (Near AI) at NeurIPS 2025'\"]},{\"title\":\"Metadata\",\"subtitle\":\"A new metadata field was added.\",\"variant\":\"added\",\"changeCount\":1,\"changes\":[\"Added metadata field 'wiki-score' with value '81'.\"]}]}]}"}