{"id":"emad-mostaque","name":"Emad Mostaque","kind":"person","aliases":[],"summary":"Emad Mostaque develops and advocates open AI systems through Intelligent Internet and contributes to Moonshots; he formerly led Stability AI; his proposals and team research retain separate attribution.","url":"https://ii.inc/blog/post/master-plan","referenceUrls":["https://ii.inc/blog/post/master-plan","https://www.ii.inc/web/blog/post/introduction-blog","https://stability.ai/news/stabilityai-announcement"],"sourceIds":["emad-mostaque-reference"],"topicIds":["intelligence-and-discovery","institutions-and-trust"],"version":"entity-catalog-v1","path":"/entities/emad-mostaque","json":"/entities/emad-mostaque.json","markdown":"/entities/emad-mostaque.md","relationships":{"sourceIds":["emad-mostaque-reference"],"noteIds":[],"perspectiveIds":[],"parallelIds":[],"topicIds":["intelligence-and-discovery","institutions-and-trust"]},"publications":"/api/pipeline/publications?entity=emad-mostaque","profile":{"schemaVersion":1,"identityId":"emad-mostaque","tracked":{"id":"emad-mostaque","name":"Emad Mostaque","role":"Intelligent Internet founder; recurring Moonshots contributor","roles":["builder","analyst"],"whyTracked":"AI-company leadership and original proposals concerning open systems and institutions.","scope":"Development, open systems and political economy; team findings retain team attribution.","interests":"Founder of Intelligent Internet; former Stability AI CEO, not its current CEO.","eligibilitySources":[{"title":"Intelligent Internet introduction","url":"https://www.ii.inc/web/blog/post/introduction-blog"},{"title":"Stability AI departure announcement","url":"https://stability.ai/news/stabilityai-announcement"}],"reviewedAt":"2026-10-14"},"contributorContexts":[],"affiliations":[{"personId":"emad-mostaque","personName":"Emad Mostaque","organizationId":"intelligent-internet","organizationName":"Intelligent Internet","role":"Intelligent Internet founder; recurring Moonshots contributor","date":"2026-10-14","dateKind":"profile-review","sources":[{"title":"Intelligent Internet introduction","url":"https://www.ii.inc/web/blog/post/introduction-blog"},{"title":"Stability AI departure announcement","url":"https://stability.ai/news/stabilityai-announcement"}]},{"personId":"emad-mostaque","personName":"Emad Mostaque","organizationId":"stability-ai","organizationName":"Stability AI","role":"Chief Executive Officer of Stability AI","date":"2023-05-16","dateKind":"source-publication","sources":[{"title":"Advocating for Open Models in AI Oversight: Stability AI's Letter to the United States Senate","url":"https://stability.ai/news-updates/stability-ai-letter-us-senate-ai-oversight"}]},{"personId":"emad-mostaque","personName":"Emad Mostaque","organizationId":"stability-ai","organizationName":"Stability AI","role":"Former Stability AI CEO, interviewed about Schelling AI","date":"2024-08-14","dateKind":"source-publication","sources":[{"title":"The Sequence Chat: Emad Mostaque -Stability AI, Schelling AI- About Open and Decentralized AI","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability"}]},{"personId":"emad-mostaque","personName":"Emad Mostaque","organizationId":"intelligent-internet","organizationName":"Intelligent Internet","role":"Founder and CEO of Intelligent Internet","date":"2025-07-24","dateKind":"source-publication","sources":[{"title":"Master Plan (open source, not so secret 🤷🏾)","url":"https://ii.inc/blog/post/master-plan"}]}],"statements":[{"id":"mostaque-sovereign-ai-2025","questionId":"power-and-risk","personId":"emad-mostaque","kind":"proposal","title":"Give people AI they own","summary":"Mostaque proposes that people own and command open, specialized agents, against a path where intelligence is centralized in opaque models and agency is rented back as a subscription.","groupId":"distributed-access","topicIds":["concentrated-power","public-ai-infrastructure"],"conceptIds":["open-models","risk-mechanisms"],"roleAtSource":"Founder and CEO of Intelligent Internet","source":{"title":"Master Plan (open source, not so secret 🤷🏾)","publisher":"Intelligent Internet","url":"https://ii.inc/blog/post/master-plan","publishedAt":"2025-07-24","dateBasis":"The page displays 24 Jul 2025 above the byline naming Emad Mostaque, Founder and CEO.","locator":"Opening diagnosis and the four-part plan, especially sovereign AI","accessNote":"Full short essay retrieved from the publisher page."},"attribution":{"kind":"author","creditedTo":"Emad Mostaque"},"conditions":["The plan proposes a benefit-backed ledger, user-owned agents, a coordination protocol, and auditable shared knowledge. Those mechanisms are proposals."],"limitations":["A manifesto. It does not show that the proposed ledger or agents exist or distribute power.","Company advocacy that promotes Intelligent Internet's own products (II-Agent, II-Medical, Common Ground)."],"support":[],"challenge":[],"evidence":[{"title":"His August 2024 interview","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability","publishedAt":"2024-08-14","relationship":"context","note":"He had already said public systems should not be black boxes controlled by an unelected few."}],"editorialAsOf":"2026-10-14","comparisonNote":"The continuity is the objection to centralized opaque control. The new piece adds an ownership plan. It is not a retraction of his scaling answer."},{"id":"mostaque-llm-scaling-unlikely-2024","questionId":"scaling-and-learning","personId":"emad-mostaque","kind":"argument","title":"AGI from scaling LLMs is unlikely","summary":"Mostaque says AGI from scaling large language models is unlikely. He compares further scaling of poor data to cooking longer, producing more capability without necessarily producing generalized knowledge in an individual model.","groupId":"additional-methods","topicIds":["scaling-laws","generalization"],"conceptIds":["scaling"],"roleAtSource":"Former Stability AI CEO, interviewed about Schelling AI","source":{"title":"The Sequence Chat: Emad Mostaque -Stability AI, Schelling AI- About Open and Decentralized AI","publisher":"The Sequence","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability","publishedAt":"2024-08-14","dateBasis":"Interview header reads Aug 14, 2024.","locator":"Rapid-fire answer to How far can the LLM scaling laws take us?","accessNote":"Original question-and-answer text read. The biography and questions are the publisher's. One earlier answer is cut off on the page and was not repaired."},"attribution":{"kind":"guest","creditedTo":"Emad Mostaque"},"conditions":["The unlikely route is AGI from scaling language models. He separately says broader systems of humans and generative models are harder to predict."],"limitations":["A short interview answer, not a technical paper.","The publisher's biography is not his claim."],"support":[],"challenge":[],"evidence":[],"editorialAsOf":"2026-10-14","comparisonNote":"This is a claim about language-model scaling, not about who should control deployed systems. That access argument is a separate answer in the same interview."},{"id":"mostaque-public-systems-not-black-boxes-2024","questionId":"power-and-risk","personId":"emad-mostaque","kind":"proposal","title":"Public systems should not be opaque or closely held","summary":"Mostaque says AI for healthcare, education and government should be open to scrutiny and not controlled by an unelected few. He sees open and closed models as complementary, while expecting closed models to outperform by adding private data and fully decentralized training to lag centralized clusters.","groupId":"distributed-access","topicIds":["concentrated-power","public-ai-infrastructure"],"conceptIds":["open-models","risk-mechanisms"],"roleAtSource":"Former Stability AI CEO, interviewed about Schelling AI","source":{"title":"The Sequence Chat: Emad Mostaque -Stability AI, Schelling AI- About Open and Decentralized AI","publisher":"The Sequence","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability","publishedAt":"2024-08-14","dateBasis":"Interview header reads Aug 14, 2024.","locator":"Answers on decentralized training, open versus closed models, and public-sector black boxes","accessNote":"Same retrieved interview. The broken sentence in the first answer was not used."},"attribution":{"kind":"guest","creditedTo":"Emad Mostaque"},"conditions":["He separates centralized training of base models from later customization and says the goal is useful distribution, not decentralization for its own sake.","He calls for new organizations to fund public AI infrastructure and expects open models to be widely used if they remain good enough."],"limitations":["Interview answers, not a statute or a system design.","The always-outperform remark rests on his reasoning that closed developers can add private data to open models, not on a benchmark."],"support":[],"challenge":[],"evidence":[{"title":"His signed Senate letter","url":"https://stability.ai/news-updates/stability-ai-letter-us-senate-ai-oversight","publishedAt":"2023-05-16","relationship":"context","note":"The letter asks for openness in the name of scrutiny and competition. It does not say closed models will always win."}],"editorialAsOf":"2026-10-14","comparisonNote":"Same day as his answer on language-model scaling. The performance and training remarks have no counterpart in his 2023 letter, so they are new topics, not a revision of it."},{"id":"mostaque-promote-openness-2023","questionId":"power-and-risk","personId":"emad-mostaque","kind":"proposal","title":"Promote open models for public scrutiny","summary":"Mostaque, signing as Stability AI's CEO, asks the Senate subcommittee to promote openness. The letter argues that open models and datasets improve safety through transparency, encourage competition and support US leadership.","groupId":"distributed-access","topicIds":["concentrated-power","policy-power-and-law","model-evaluation-rules"],"conceptIds":["open-models","independent-evaluation"],"roleAtSource":"Chief Executive Officer of Stability AI","source":{"title":"Advocating for Open Models in AI Oversight: Stability AI's Letter to the United States Senate","publisher":"Stability AI","url":"https://stability.ai/news-updates/stability-ai-letter-us-senate-ai-oversight","publishedAt":"2023-05-16","dateBasis":"The announcement page's datePublished metadata is 2023-05-16T16:48:22+0000. The signed letter itself is dated May 13, 2023. The public release date is used.","locator":"Signed cover letter, PDF pages 1–2","accessNote":"The signed letter and attached company submission were read from the publisher PDF. The announcement page's dateModified is 2024-10-14, so the current page is not assumed to be an untouched May 2023 snapshot."},"attribution":{"kind":"author","creditedTo":"Emad Mostaque, signing for Stability AI"},"conditions":["The letter ties openness to public scrutiny, competition and US leadership, and says public oversight starts with public scrutiny.","It cites his signature on the March 2023 open letter urging caution and Stability AI's participation in community-led model testing announced by the White House."],"limitations":["Written mostly in Stability AI's voice and signed by Mostaque as CEO. The attached risk taxonomy and five oversight suggestions are the company's submission.","The announcement page names a different CMS author and was modified on 14 October 2024; the signed letter itself is dated 13 May 2023."],"support":[],"challenge":[],"evidence":[],"editorialAsOf":"2026-10-14","comparisonNote":"This is his signed request for openness. It does not say that every model must be trained in a decentralized way or that open models will match closed ones."}],"questions":[{"id":"scaling-and-learning","title":"Can AI keep improving without more of the same data?","shortTitle":"Scaling & learning","summary":"A limit on public text is not a limit on all learning. The dispute concerns what generalizes, what needs interaction and which feedback can actually be obtained.","topicIds":["ai-capabilities-and-limits","training-data-availability","scaling-laws"],"conceptIds":["scaling","continual-learning","verifiable-feedback"],"groups":[{"id":"additional-methods","label":"Additional methods needed","description":"The existing approach may need different learning mechanisms or representations."},{"id":"scale-and-generalize","label":"Scale and generalize","description":"Broader training and compute may produce the required general capabilities."},{"id":"learn-from-experience","label":"Learn from experience","description":"Scarce real-world feedback could matter more than replaying familiar tasks."}],"editorialAsOf":"2026-10-14","coverageNote":"Selected research and statements from 2016–2026, including joint experiments and dated hearings. Research findings retain their coauthors and experimental limits; they are not interchangeable AGI forecasts."},{"id":"power-and-risk","title":"Who should have power over increasingly capable AI?","shortTitle":"Power & risk","summary":"Access, independent scrutiny and protection against severe harms can pull in different directions. Each argument needs its own mechanism and evidence.","topicIds":["policy-power-and-law","security-safety-and-resilience","concentrated-power"],"conceptIds":["open-models","risk-mechanisms","independent-evaluation"],"groups":[{"id":"distributed-access","label":"Distribute access","description":"Concentrated control can itself create harm."},{"id":"independent-scrutiny","label":"When and how to scrutinize","description":"Arguments about who evaluates AI and whether checks should happen before release or throughout use."},{"id":"extreme-risk-preparation","label":"Prepare for severe harms","description":"Technical safeguards and governance may need to advance together."}],"editorialAsOf":"2026-10-14","coverageNote":"Selected 2022–2026 proposals and arguments. The groups can overlap; they do not assign a general optimism or doom score to a person."}],"readings":[],"relations":[{"id":"mostaque-promote-openness-2023:mostaque-public-systems-not-black-boxes-2024","kind":"continued","question":"Who should have power over increasingly capable AI?","topicIds":["concentrated-power","policy-power-and-law","model-evaluation-rules","public-ai-infrastructure"],"summary":"Both call for AI that is open to public scrutiny. The 2023 letter says open models improve safety through transparency; the 2024 interview says public-sector AI should not run on black boxes controlled by an unelected few.","boundary":"Continuity is the transparency and scrutiny claim. The 2023 letter is a company submission he signed as CEO; the 2024 answers are personal, after leaving Stability AI. His 2024 remarks that closed models will outperform and decentralized training will lag have no 2023 counterpart, so they are not a revision.","editorialAsOf":"2026-10-14","basis":"explicit-editorial-assessment","records":[{"kind":"statement","id":"mostaque-promote-openness-2023","title":"Promote open models for public scrutiny","entityId":"emad-mostaque","date":"2023-05-16","url":"https://stability.ai/news-updates/stability-ai-letter-us-senate-ai-oversight","path":"/tracker/?question=power-and-risk&statement=mostaque-promote-openness-2023"},{"kind":"statement","id":"mostaque-public-systems-not-black-boxes-2024","title":"Public systems should not be opaque or closely held","entityId":"emad-mostaque","date":"2024-08-14","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability","path":"/tracker/?question=power-and-risk&statement=mostaque-public-systems-not-black-boxes-2024"}]},{"id":"mostaque-public-systems-not-black-boxes-2024:mostaque-sovereign-ai-2025","kind":"continued","question":"Who should have power over increasingly capable AI?","topicIds":["concentrated-power","public-ai-infrastructure"],"summary":"Both reject centralized opaque intelligence for systems people depend on. The later plan proposes agents that users own and command.","boundary":"Continuity is that control objection. The July 2025 plan does not restate, and therefore does not confirm, the August 2024 claim that closed models will always outperform open ones.","editorialAsOf":"2026-10-14","basis":"explicit-editorial-assessment","records":[{"kind":"statement","id":"mostaque-public-systems-not-black-boxes-2024","title":"Public systems should not be opaque or closely held","entityId":"emad-mostaque","date":"2024-08-14","url":"https://thesequence.substack.com/p/the-sequence-chat-emad-mostaque-stability","path":"/tracker/?question=power-and-risk&statement=mostaque-public-systems-not-black-boxes-2024"},{"kind":"statement","id":"mostaque-sovereign-ai-2025","title":"Give people AI they own","entityId":"emad-mostaque","date":"2025-07-24","url":"https://ii.inc/blog/post/master-plan","path":"/tracker/?question=power-and-risk&statement=mostaque-sovereign-ai-2025"}]}],"boundary":"Escape Meridian selects and paraphrases these records. A mention, shared topic or affiliation does not establish authorship, endorsement, agreement or causation. Missing positions remain unknown."},"currentReadings":{"entityId":"emad-mostaque","items":[],"scanned":12,"truncated":false,"scanLimit":500,"checkedAt":"2026-10-18T03:58:50.380Z"},"currentReadingsUrl":"/entities/emad-mostaque.json"}