> ## Documentation Index
> Fetch the complete documentation index at: https://usefulai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Best AI LinkedIn Accounts to Follow in 2026

export const individuals = [
  {
    "slug": "satya-nadella-linkedin",
    "name": "Satya Nadella",
    "url": "https://www.linkedin.com/in/satyanadella/",
    "image": "/images/feeds/linkedin/satya-nadella-linkedin.png",
    "bestFor": "Microsoft's AI strategy and impact",
    "metric": "12M followers",
    "facts": [
      "12M followers",
      "Microsoft chairman and CEO"
    ],
    "about": "Satya Nadella uses LinkedIn to present Microsoft's AI strategy through major product, infrastructure, research, customer, and public-interest announcements.",
    "take": "The account is a primary executive source for Microsoft's direction and the partnerships behind it. Its scale and corporate scope make it useful for strategy, not for independent product evaluation.",
    "highlights": [
      {
        "title": "Copilot Cowork General Availability Multi-Model Support",
        "url": "https://www.linkedin.com/posts/satyanadella_copilot-cowork-is-now-generally-available-activity-7472669034983936003-ePub",
        "date": "2026-06-16",
        "meta": "4,967 likes"
      },
      {
        "title": "Jensen Huang Satya Nadella Microsoft Build Conversation",
        "url": "https://www.linkedin.com/posts/satyanadella_jensen-huang-and-satya-nadellas-conversation-activity-7468018920784875522-UoS5",
        "date": "2026-06-03",
        "meta": "1,836 likes"
      },
      {
        "title": "AI Research on Cell Behavior and Cancer Medicine Response",
        "url": "https://www.linkedin.com/posts/satyanadella_why-dont-cancer-medicines-work-the-same-activity-7470160151099006976-w-EO",
        "date": "2026-06-09",
        "meta": "1,249 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "andrew-ng-linkedin",
    "name": "Andrew Ng",
    "url": "https://www.linkedin.com/in/andrewyng/",
    "image": "/images/feeds/linkedin/andrew-ng-linkedin.png",
    "bestFor": "Practical AI education and adoption",
    "metric": "3M followers",
    "facts": [
      "3M followers",
      "AI educator and entrepreneur"
    ],
    "about": "Andrew Ng combines the perspective of a machine-learning researcher, educator, and company builder, using LinkedIn to share courses, applied projects, and views on how AI changes work.",
    "take": "The account is strongest for grounded educational opportunities and adoption lessons rather than launch-by-launch commentary. His own companies and courses appear frequently, but the posts usually explain the underlying skill or market change clearly.",
    "highlights": [
      {
        "title": "AI Forward Deployed Engineers and Job Market Growth",
        "url": "https://www.linkedin.com/posts/andrewyng_one-of-the-new-buzzy-jobs-in-silicon-valley-activity-7467243316431036416-fzyJ",
        "date": "2026-06-01",
        "meta": "14,385 likes"
      },
      {
        "title": "Efficient LLM Serving Course with Red Hat vLLM",
        "url": "https://www.linkedin.com/posts/andrewyng_new-course-on-serving-llms-efficiently-activity-7468342916764250112-8HxO",
        "date": "2026-06-04",
        "meta": "3,521 likes"
      },
      {
        "title": "Bob Bradway on AI Transformation at Amgen",
        "url": "https://www.linkedin.com/posts/andrewyng_ai-and-the-future-of-pharma-with-bob-bradway-activity-7474962537612984320-UJ_j",
        "date": "2026-06-22",
        "meta": "756 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "yann-lecun-linkedin",
    "name": "Yann LeCun",
    "url": "https://www.linkedin.com/in/yann-lecun/",
    "image": "/images/feeds/linkedin/yann-lecun-linkedin.png",
    "bestFor": "Open research and alternatives to LLM scaling",
    "metric": "1M followers",
    "facts": [
      "1M followers",
      "AI researcher and professor"
    ],
    "about": "Yann LeCun publishes research, commentary, and debate about world models, self-supervised learning, open AI, and the limits of scaling current language-model approaches.",
    "take": "The account is high-signal precisely because it offers a technically informed counterposition to dominant frontier-lab narratives. The volume and argumentative style require separating research evidence from opinion, but the perspective is essential.",
    "highlights": [
      {
        "title": "AI Labor Market Effects and IPO Timing Concerns",
        "url": "https://www.linkedin.com/posts/yann-lecun_ive-said-to-not-listen-to-sam-and-dario-activity-7465173471610863617-GVrU",
        "date": "2026-05-26",
        "meta": "8,676 likes"
      },
      {
        "title": "US Government Bans Mythos/Fable Model for Non-Americans",
        "url": "https://www.linkedin.com/posts/yann-lecun_dario-amodeis-ridiculous-fear-mongering-activity-7471505011575984128-95aK",
        "date": "2026-06-13",
        "meta": "6,683 likes"
      },
      {
        "title": "AI Doomerism and Apocalyptic Cult Characteristics",
        "url": "https://www.linkedin.com/posts/yann-lecun_ai-doomerism-has-all-the-characteristics-activity-7467551332300701698-L1fB",
        "date": "2026-06-02",
        "meta": "2,195 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "kai-fu-lee-linkedin",
    "name": "Kai-Fu Lee",
    "url": "https://www.linkedin.com/in/kaifulee/",
    "image": "/images/feeds/linkedin/kai-fu-lee-linkedin.png",
    "bestFor": "Global AI strategy",
    "metric": "889K followers",
    "facts": [
      "889K followers",
      "AI investor and computer scientist"
    ],
    "about": "Kai-Fu Lee brings a cross-border research, company-building, and investment perspective to questions about AI development in China, the United States, and other national ecosystems.",
    "take": "The account is most valuable for strategic comparisons that are underrepresented in US-centric feeds. Posting is relatively sparse, but the perspective is distinct and usually tied to larger talks or arguments.",
    "highlights": [
      {
        "title": "AI Sovereignty: Three Strategic Paths for Countries",
        "url": "https://www.linkedin.com/posts/kaifulee_my-latest-interview-with-capgemini-sharing-activity-7459495375935102976-2k68",
        "date": "2026-05-11",
        "meta": "252 likes"
      },
      {
        "title": "U.S. and China AI Development Cycles and Openness Shift",
        "url": "https://www.linkedin.com/posts/kaifulee_us-and-china-ai-development-will-continue-activity-7470567580022304768-5TL1",
        "date": "2026-06-10",
        "meta": "118 likes"
      },
      {
        "title": "Kai-Fu Lee speaks at Asia House London",
        "url": "https://www.linkedin.com/posts/kaifulee_its-a-privilege-to-speak-at-the-asia-house-activity-7470326854017691649-BorS",
        "date": "2026-06-10",
        "meta": "73 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "aravind-srinivas-linkedin",
    "name": "Aravind Srinivas",
    "url": "https://www.linkedin.com/in/aravind-srinivas-16051987/",
    "image": "/images/feeds/linkedin/aravind-srinivas-linkedin.png",
    "bestFor": "AI search and product building",
    "metric": "866K followers",
    "facts": [
      "866K followers",
      "Perplexity co-founder and CEO"
    ],
    "about": "Aravind Srinivas uses LinkedIn as a founder-operator account for Perplexity, mixing product updates with observations about AI search, agents, and how knowledge work may change.",
    "take": "Follow him for first-hand Perplexity direction and the product reasoning behind launches. The account is naturally promotional and optimistic, so broader market or labor claims need independent evidence.",
    "highlights": [
      {
        "title": "Perplexity Intel Ultra Series 3 Local AI Models",
        "url": "https://www.linkedin.com/posts/aravind-srinivas-16051987_thank-you-lip-bu-tan-and-intel-for-working-activity-7469107780302393344-VlK1",
        "date": "2026-06-06",
        "meta": "2,952 likes"
      },
      {
        "title": "Harvard Study: AI Agents Reshape Knowledge Work Efficiency",
        "url": "https://www.linkedin.com/posts/aravind-srinivas-16051987_together-with-harvard-business-school-were-activity-7469794162674929664-ZYg0",
        "date": "2026-06-08",
        "meta": "522 likes"
      },
      {
        "title": "Podcast with Harry Stebbings on Export Controls and AI",
        "url": "https://www.linkedin.com/posts/aravind-srinivas-16051987_how-export-controls-helped-not-hurt-china-activity-7472446038495285248-Wn2q",
        "date": "2026-06-16"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "ruben-hassid-linkedin",
    "name": "Ruben Hassid",
    "url": "https://www.linkedin.com/in/ruben-hassid/",
    "image": "/images/feeds/linkedin/ruben-hassid-linkedin.png",
    "bestFor": "Practical Claude and prompting education",
    "metric": "811.7K followers",
    "facts": [
      "811.7K followers",
      "AI creator and entrepreneur"
    ],
    "about": "Ruben Hassid publishes frequent, approachable guides for using Claude and other AI products, often packaging techniques into reusable prompts and learning resources.",
    "take": "The account is useful for nontechnical users who want concrete examples and a clear starting point. Its high cadence and strong hooks make it a discovery feed; more complex claims should be tested before becoming a standard workflow.",
    "highlights": [
      {
        "title": "Free Claude Mastery Guides Four Levels",
        "url": "https://www.linkedin.com/posts/ruben-hassid_instead-of-watching-netflix-tonight-spend-activity-7474053489728430080-TCBU",
        "date": "2026-06-20",
        "meta": "1,566 likes"
      },
      {
        "title": "Claude 5 Fable AI Model Features and Usage",
        "url": "https://www.linkedin.com/posts/ruben-hassid_the-new-claude-5-is-the-smartest-ai-you-ever-activity-7470429572413677570-8MIt",
        "date": "2026-06-10",
        "meta": "1,074 likes"
      },
      {
        "title": "Claude Fable 5 Prompt Anatomy Best Practices",
        "url": "https://www.linkedin.com/posts/ruben-hassid_the-anatomy-of-a-new-claude-fable-5-prompt-activity-7470791976456667136-Aa26",
        "date": "2026-06-11",
        "meta": "1,114 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "jensen-huang-linkedin",
    "name": "Jensen Huang",
    "url": "https://www.linkedin.com/in/jenhsunhuang/",
    "image": "/images/feeds/linkedin/jensen-huang-linkedin.png",
    "bestFor": "AI computing infrastructure",
    "metric": "757K followers",
    "facts": [
      "757K followers",
      "NVIDIA founder and CEO"
    ],
    "about": "Jensen Huang's LinkedIn account presents NVIDIA's view of AI as a new computing and industrial infrastructure layer, spanning chips, systems, software, and robotics.",
    "take": "The account is useful for tracking NVIDIA's strategic narrative and major ecosystem announcements from its CEO. It is not an independent guide to infrastructure economics or competitive performance.",
    "highlights": [
      {
        "title": "NVIDIA Computex Announcements Vera Rubin RTX Spark",
        "url": "https://www.linkedin.com/posts/jenhsunhuang_today-at-computex-in-taipei-i-had-the-opportunity-activity-7467202810439200768-7dTV",
        "date": "2026-06-01",
        "meta": "15,771 likes"
      },
      {
        "title": "AI as Essential Infrastructure Five-Layer Stack",
        "url": "https://www.linkedin.com/posts/jenhsunhuang_ai-is-one-of-the-most-powerful-forces-shaping-activity-7437075028833247232-hBKm",
        "date": "2026-03-10",
        "meta": "8,894 likes"
      },
      {
        "title": "NVIDIA Blackwell Chip Manufacturing Begins in United States",
        "url": "https://www.linkedin.com/posts/jenhsunhuang_today-marks-a-new-chapter-for-american-manufacturing-activity-7385125951698866176-iiwH",
        "date": "2025-10-18",
        "meta": "4,836 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "joerg-storm-linkedin",
    "name": "Dr. Joerg Storm",
    "url": "https://www.linkedin.com/in/joergstorm/",
    "image": "/images/feeds/linkedin/joerg-storm-linkedin.png",
    "bestFor": "AI adoption for business professionals",
    "metric": "705.3K followers",
    "facts": [
      "705.3K followers",
      "AI business creator and newsletter author"
    ],
    "about": "Dr. Joerg Storm publishes high-frequency business-facing AI commentary and the DIGITAL STORM weekly newsletter, translating product changes into workplace and management implications.",
    "take": "The account is useful for managers who want AI developments framed as operational changes rather than technical research. The large, frequent output and strong hooks create repetition, so prioritize the deeper weekly articles over every short post.",
    "highlights": [
      {
        "title": "AI Escaped the Feed Future of Work",
        "url": "https://www.linkedin.com/posts/joergstorm_ai-artificialintelligence-futureofwork-activity-7472972868285337601-6YRD",
        "date": "2026-06-17",
        "meta": "4,952 likes"
      },
      {
        "title": "Viktor AI Employee Microsoft Teams Launch",
        "url": "https://www.linkedin.com/posts/joergstorm_%F0%9D%97%A0%F0%9D%97%BC%F0%9D%98%80%F0%9D%98%81-%F0%9D%97%94%F0%9D%97%9C-%F0%9D%97%AE%F0%9D%97%BB%F0%9D%98%80%F0%9D%98%84%F0%9D%97%B2%F0%9D%97%BF%F0%9D%98%80-%F0%9D%97%A9%F0%9D%97%B6%F0%9D%97%B8%F0%9D%98%81%F0%9D%97%BC%F0%9D%97%BF-activity-7473365034925105153-ZM14",
        "date": "2026-06-18",
        "meta": "335 likes"
      },
      {
        "title": "AI Strategy Generation Versus Consultant Costs",
        "url": "https://www.linkedin.com/posts/joergstorm_ai-aiagents-entrepreneurship-activity-7473343841673289730-5F6-",
        "date": "2026-06-18",
        "meta": "324 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "cassie-kozyrkov-linkedin",
    "name": "Cassie Kozyrkov",
    "url": "https://www.linkedin.com/in/kozyrkov/",
    "image": "/images/feeds/linkedin/cassie-kozyrkov-linkedin.png",
    "bestFor": "Better decisions about AI",
    "metric": "698K followers",
    "facts": [
      "698K followers",
      "Decision intelligence leader and educator"
    ],
    "about": "Cassie Kozyrkov explains AI through the lens of decision quality, helping leaders separate what a model can produce from the judgment required to use it well.",
    "take": "Her account is valuable because it challenges sloppy reasoning around AI adoption instead of only reviewing tools. The strongest posts provide frameworks that remain useful after a specific model or feature changes.",
    "highlights": [
      {
        "title": "Employers Regret AI-Driven Layoffs",
        "url": "https://www.linkedin.com/posts/kozyrkov_aafa-decisionintelligence-aileadership-activity-7472256695671640064-9x72",
        "date": "2026-06-15",
        "meta": "125 likes"
      },
      {
        "title": "Three Organizational Design Changes for AI Era",
        "url": "https://www.linkedin.com/posts/kozyrkov_aafa-decisionintelligence-aileadership-activity-7477330042381860864-T-sC",
        "date": "2026-06-29",
        "meta": "85 likes"
      },
      {
        "title": "LLM Decided Is Not An Explanation",
        "url": "https://www.linkedin.com/posts/kozyrkov_decisionintelligence-aileadership-governance-activity-7470807248080900096-qSFn",
        "date": "2026-06-11",
        "meta": "82 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "mustafa-suleyman-linkedin",
    "name": "Mustafa Suleyman",
    "url": "https://www.linkedin.com/in/mustafa-suleyman/",
    "image": "/images/feeds/linkedin/mustafa-suleyman-linkedin.png",
    "bestFor": "Consumer AI and organizational strategy",
    "metric": "534K followers",
    "facts": [
      "534K followers",
      "Microsoft AI CEO"
    ],
    "about": "Mustafa Suleyman shares executive-level views from Microsoft AI about building consumer AI products, organizing high-talent teams, and pursuing advanced systems responsibly.",
    "take": "The account is useful for understanding the priorities and culture of Microsoft's dedicated AI organization. It is selective and aspirational, with less operational or technical detail than practitioner feeds.",
    "highlights": [
      {
        "title": "Microsoft AI Team Culture Principles",
        "url": "https://www.linkedin.com/posts/activity-7476291606841540608-gQiA",
        "date": "2026-06-26",
        "meta": "492 likes"
      },
      {
        "title": "Talent Density in Microsoft AI Humanist Superintelligence",
        "url": "https://www.linkedin.com/posts/activity-7473394178161319937-WIO8",
        "date": "2026-06-18",
        "meta": "200 likes"
      },
      {
        "title": "Superintelligence Team Meetup Boston",
        "url": "https://www.linkedin.com/posts/activity-7475240917512187905--KWf",
        "date": "2026-06-23",
        "meta": "198 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "ethan-mollick-linkedin",
    "name": "Ethan Mollick",
    "url": "https://www.linkedin.com/in/emollick/",
    "image": "/images/feeds/linkedin/ethan-mollick-linkedin.png",
    "bestFor": "Evidence-based AI use at work and school",
    "metric": "398K followers",
    "facts": [
      "398K followers",
      "Wharton professor and AI researcher"
    ],
    "about": "Ethan Mollick translates research and sustained model experimentation into practical observations about work, education, management, and how people adapt to AI.",
    "take": "This is one of the highest-signal individual LinkedIn feeds because claims are often tied to a paper, experiment, or clearly described test. The volume is high, but the posts add interpretation rather than merely repeating announcements.",
    "highlights": [
      {
        "title": "Recognizing AI Generated Content Online",
        "url": "https://www.linkedin.com/posts/emollick_as-more-people-come-to-recognize-the-tells-activity-7464445097716375552-vj_G",
        "date": "2026-05-24",
        "meta": "3,288 likes"
      },
      {
        "title": "Co-Existence Book October 20 Release",
        "url": "https://www.linkedin.com/posts/emollick_i-have-a-new-book-coming-out-october-20-activity-7468411654381289472-oxlx",
        "date": "2026-06-04",
        "meta": "3,223 likes"
      },
      {
        "title": "Gemini Omni Native Video Editing Capabilities",
        "url": "https://www.linkedin.com/posts/emollick_i-think-people-dont-realize-why-gemini-omni-activity-7463643192136196096-NKlr",
        "date": "2026-05-22",
        "meta": "1,052 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "clem-delangue-linkedin",
    "name": "Clem Delangue",
    "url": "https://www.linkedin.com/in/clementdelangue/",
    "image": "/images/feeds/linkedin/clem-delangue-linkedin.png",
    "bestFor": "Open AI ecosystem strategy",
    "metric": "305K followers",
    "facts": [
      "305K followers",
      "Hugging Face co-founder and CEO"
    ],
    "about": "Clem Delangue shares a founder's view of the open AI ecosystem, including Hugging Face platform growth, local model adoption, hardware partnerships, and open-source policy.",
    "take": "This is one of the more useful executive accounts for tracking open-model infrastructure and adoption signals. Company milestones appear frequently, but the posts often surface concrete ecosystem data that is difficult to find elsewhere.",
    "highlights": [
      {
        "title": "Stanford Study Shows Local AI Models Answer 71% of Queries",
        "url": "https://www.linkedin.com/posts/clementdelangue_a-study-from-stanford-university-showed-that-activity-7477717394812182529-eRtA",
        "date": "2026-06-30",
        "meta": "1,099 likes"
      },
      {
        "title": "On-Prem Local AI with Hugging Face and Dell",
        "url": "https://www.linkedin.com/posts/clementdelangue_i-believe-on-prem-and-local-ai-based-on-activity-7462205366752399360-ygQv",
        "date": "2026-05-18",
        "meta": "609 likes"
      },
      {
        "title": "AMD Ryzen AI Halo Local Hardware for AI Builders",
        "url": "https://www.linkedin.com/posts/clementdelangue_im-excited-about-the-new-amd-ryzen-ai-halo-activity-7463340648633516033-_p45",
        "date": "2026-05-21",
        "meta": "330 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "demis-hassabis-linkedin",
    "name": "Demis Hassabis",
    "url": "https://www.linkedin.com/in/demishassabis/",
    "image": "/images/feeds/linkedin/demis-hassabis-linkedin.png",
    "bestFor": "AI research and scientific impact",
    "metric": "293K followers",
    "facts": [
      "293K followers",
      "Google DeepMind co-founder and CEO"
    ],
    "about": "Demis Hassabis uses LinkedIn to connect Google DeepMind's model work with its larger scientific agenda, including biology, health, and international research partnerships.",
    "take": "The account is strongest for first-hand statements about DeepMind's research direction and scientific ambitions. Posts are selective and executive-level, so technical details usually live in the linked research.",
    "highlights": [
      {
        "title": "AI for Human Health AlphaFold Isomorphic Labs Funding",
        "url": "https://www.linkedin.com/posts/demishassabis_ive-always-believed-the-no1-application-activity-7459967883570384896-xEUx",
        "date": "2026-05-12",
        "meta": "14,829 likes"
      },
      {
        "title": "Google I/O: Gemini Omni, Flash, and AI Safety",
        "url": "https://www.linkedin.com/posts/demishassabis_it-was-great-to-be-at-io-again-this-year-activity-7463305686215557120-7TLQ",
        "date": "2026-05-21",
        "meta": "5,576 likes"
      },
      {
        "title": "AlphaGo 10th Anniversary Korea Visit and AI Partnership",
        "url": "https://www.linkedin.com/posts/demishassabis_it-was-wonderful-to-be-back-in-korea-last-activity-7459742448072867840-xWAn",
        "date": "2026-05-11",
        "meta": "3,907 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "fei-fei-li-linkedin",
    "name": "Fei-Fei Li",
    "url": "https://www.linkedin.com/in/fei-fei-li-4541247/",
    "image": "/images/feeds/linkedin/fei-fei-li-linkedin.png",
    "bestFor": "Human-centered and spatial AI",
    "metric": "127K followers",
    "facts": [
      "127K followers",
      "AI researcher and Stanford professor"
    ],
    "about": "Fei-Fei Li shares research and company work around spatial intelligence and world models while maintaining a consistent human-centered view of AI's purpose.",
    "take": "The account is especially useful for following the shift beyond language toward visual and spatial intelligence from a leading researcher. Posts are selective and often point to substantive papers or demonstrations.",
    "highlights": [
      {
        "title": "AI Should Serve People, Not Reverse",
        "url": "https://www.linkedin.com/posts/fei-fei-li-4541247_ive-spent-my-career-trying-to-make-sure-activity-7473109039145463808-fYDK",
        "date": "2026-06-17",
        "meta": "5,342 likes"
      },
      {
        "title": "Spatial Intelligence as AI's Next Frontier Beyond Language",
        "url": "https://www.linkedin.com/posts/fei-fei-li-4541247_from-words-to-worlds-spatial-intelligence-activity-7393658794339295232-Qdd9",
        "date": "2025-11-10",
        "meta": "4,936 likes"
      },
      {
        "title": "World Model Taxonomy in AI Research",
        "url": "https://www.linkedin.com/posts/fei-fei-li-4541247_world-model-has-become-one-of-the-most-activity-7468001018635554816-uS-b",
        "date": "2026-06-03",
        "meta": "4,706 likes"
      }
    ],
    "platforms": [

    ]
  }
];

export const companies = [
  {
    "slug": "openai-linkedin",
    "name": "OpenAI",
    "url": "https://www.linkedin.com/company/openai/",
    "image": "/images/feeds/linkedin/openai-linkedin.png",
    "bestFor": "OpenAI company and product updates",
    "metric": "10.87M followers",
    "facts": [
      "10.87M followers",
      "Official OpenAI account"
    ],
    "about": "OpenAI's LinkedIn feed is a broad first-party stream for model and product releases, research, infrastructure, partnerships, and corporate milestones.",
    "take": "This is essential for confirming OpenAI announcements and seeing how the company frames new products. It is also the largest company account in the set, which makes raw follower ordering a poor substitute for editorial relevance.",
    "highlights": [
      {
        "title": "OpenAI Designs and Builds Jalapeño AI Chip",
        "url": "https://www.linkedin.com/posts/openai_weve-designed-and-built-our-first-ai-chip-activity-7475540009161248768-s7RI",
        "date": "2026-06-24",
        "meta": "3,288 likes"
      },
      {
        "title": "Codex Preview in ChatGPT Mobile App",
        "url": "https://www.linkedin.com/posts/openai_work-with-codex-from-anywhere-activity-7460784183288971264-BkYW",
        "date": "2026-05-14",
        "meta": "2,852 likes"
      },
      {
        "title": "GPT-5.6 Sol, Terra, Luna Limited Preview",
        "url": "https://www.linkedin.com/posts/openai_previewing-gpt-56-sol-a-next-generation-activity-7476321920909762560-u8y4",
        "date": "2026-06-26",
        "meta": "2,000 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "anthropic-linkedin",
    "name": "Anthropic",
    "url": "https://www.linkedin.com/company/anthropicresearch/",
    "image": "/images/feeds/linkedin/anthropic-linkedin.png",
    "bestFor": "Anthropic company and product updates",
    "metric": "3.25M followers",
    "facts": [
      "3.25M followers",
      "Official Anthropic account"
    ],
    "about": "Anthropic's LinkedIn account is a first-party stream for Claude releases, research, safety positions, partnerships, hiring, and company news.",
    "take": "This is a direct source for material developments at Anthropic and often adds useful business context beyond a release note. It should not be used alone for model comparisons or policy questions where the company has a clear stake.",
    "highlights": [
      {
        "title": "Anthropic SpaceX Partnership Increases Compute Capacity",
        "url": "https://www.linkedin.com/posts/anthropicresearch_higher-usage-limits-for-claude-and-a-compute-activity-7457829061063688192-TKup",
        "date": "2026-05-06",
        "meta": "14,245 likes"
      },
      {
        "title": "Anthropic Confidentially Submits S-1 Registration Statement",
        "url": "https://www.linkedin.com/posts/anthropicresearch_anthropic-confidentially-submits-draft-s-activity-7467244047657635840--SrW",
        "date": "2026-06-01",
        "meta": "10,683 likes"
      },
      {
        "title": "US Government Export Control Suspends Fable 5 Mythos 5",
        "url": "https://www.linkedin.com/posts/anthropicresearch_statement-on-the-us-government-directive-activity-7471375068825010176--x6_",
        "date": "2026-06-13",
        "meta": "6,991 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "perplexity-ai-linkedin",
    "name": "Perplexity",
    "url": "https://www.linkedin.com/company/perplexity-ai/",
    "image": "/images/feeds/linkedin/perplexity-ai-linkedin.png",
    "bestFor": "AI search and agent products",
    "metric": "1.63M followers",
    "facts": [
      "1.63M followers",
      "Official Perplexity account"
    ],
    "about": "Perplexity's LinkedIn feed publishes rapid first-party product updates across search, research, browser, and agent functionality.",
    "take": "The account is useful for staying current with a product that changes quickly and integrates many outside models. Repeated launch posts can exaggerate novelty, so evaluate whether a feature changes the actual research workflow.",
    "highlights": [
      {
        "title": "Claude Fable 5 Available in Computer as Orchestrator",
        "url": "https://www.linkedin.com/posts/perplexity-ai_claude-fable-5-is-now-available-in-computer-activity-7470537116758769664-urtb",
        "date": "2026-06-10",
        "meta": "688 likes"
      },
      {
        "title": "Claude Fable 5 Available as Orchestrator Model",
        "url": "https://www.linkedin.com/posts/perplexity-ai_claude-fable-5-is-once-again-available-in-activity-7478198826365906944-r42a",
        "date": "2026-07-01",
        "meta": "222 likes"
      },
      {
        "title": "Deep Research Integrated as Native Computer Skill",
        "url": "https://www.linkedin.com/posts/perplexity-ai_were-integrating-deep-research-as-a-native-activity-7470890712683720704-hldO",
        "date": "2026-06-11",
        "meta": "87 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "deepmind-linkedin",
    "name": "Google DeepMind",
    "url": "https://www.linkedin.com/company/googledeepmind/",
    "image": "/images/feeds/linkedin/deepmind-linkedin.png",
    "bestFor": "Frontier AI research and products",
    "metric": "1.56M followers",
    "facts": [
      "1.56M followers",
      "Official Google DeepMind account"
    ],
    "about": "Google DeepMind's LinkedIn account combines model and product announcements with research on science, agents, robotics, and responsible AI.",
    "take": "The feed is valuable as a primary source for one of the most consequential AI labs and gives research applications more space than most company accounts. It still presents the lab's own framing, particularly around capability and safety.",
    "highlights": [
      {
        "title": "Gemini 3.5 Flash Model Release",
        "url": "https://www.linkedin.com/posts/googledeepmind_gemini-35-is-our-newest-family-of-models-activity-7462553838139502592-vFln",
        "date": "2026-05-19",
        "meta": "6,446 likes"
      },
      {
        "title": "Gemini Omni: Character and Scene Consistency",
        "url": "https://www.linkedin.com/posts/googledeepmind_introducing-gemini-omni-activity-7462605788314894336-MFNL",
        "date": "2026-05-19",
        "meta": "3,613 likes"
      },
      {
        "title": "Gemini for Science: Three Agentic Prototypes",
        "url": "https://www.linkedin.com/posts/googledeepmind_were-introducing-gemini-for-science-a-collection-activity-7462865515753967618-LHHp",
        "date": "2026-05-20",
        "meta": "2,936 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "huggingface-linkedin",
    "name": "Hugging Face",
    "url": "https://www.linkedin.com/company/huggingface/",
    "image": "/images/feeds/linkedin/huggingface-linkedin.png",
    "bestFor": "Open AI infrastructure",
    "metric": "1.24M followers",
    "facts": [
      "1.24M followers",
      "Official Hugging Face account"
    ],
    "about": "Hugging Face's LinkedIn account publishes platform and ecosystem updates across models, datasets, hardware, open science, and community growth.",
    "take": "This is a useful source for concrete open-AI ecosystem milestones and product changes. The account covers a very broad platform, so technical users will often need the linked repository or documentation for actionable detail.",
    "highlights": [
      {
        "title": "Google Releases Gemma 4 Apache 2.0 Local",
        "url": "https://www.linkedin.com/posts/huggingface_so-happy-to-see-google-release-gemma-4-today-activity-7445518953755389952-3oI4",
        "date": "2026-04-02",
        "meta": "4,132 likes"
      },
      {
        "title": "Hugging Face Hardware Community Usage Report",
        "url": "https://www.linkedin.com/posts/huggingface_introducing-hf-hardware-activity-7462938585676038145-lv2P",
        "date": "2026-05-20",
        "meta": "602 likes"
      },
      {
        "title": "Hugging Face Hub Reaches 1 Million Open Datasets",
        "url": "https://www.linkedin.com/posts/huggingface_weve-just-hit-1m-open-datasets-on-the-hugging-activity-7459989642499776512-W82G",
        "date": "2026-05-12",
        "meta": "230 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "mistral-linkedin",
    "name": "Mistral AI",
    "url": "https://www.linkedin.com/company/mistralai/",
    "image": "/images/feeds/linkedin/mistral-linkedin.png",
    "bestFor": "Mistral models and enterprise AI",
    "metric": "643.1K followers",
    "facts": [
      "643.1K followers",
      "Official Mistral AI account"
    ],
    "about": "Mistral AI's LinkedIn feed combines first-party model and product news with partnerships, acquisitions, events, and the company's role in the European AI ecosystem.",
    "take": "This is a useful source for understanding Mistral as both a model developer and enterprise platform company. Product and corporate news are mixed together, so technical readers should follow the linked documentation.",
    "highlights": [
      {
        "title": "Mistral AI Now Summit Paris 1500 Attendees",
        "url": "https://www.linkedin.com/posts/mistralai_thank-you-to-the-over-1500-attendees-who-activity-7466115400246738944-JQ3s",
        "date": "2026-05-29",
        "meta": "3,998 likes"
      },
      {
        "title": "Mistral AI SAP Business AI Platform Partnership",
        "url": "https://www.linkedin.com/posts/mistralai_were-thrilled-to-announce-a-new-mistral-activity-7463064994637754368-LgJf",
        "date": "2026-05-21",
        "meta": "3,427 likes"
      },
      {
        "title": "Mistral AI Acquires Emmi AI Physics Engineering",
        "url": "https://www.linkedin.com/posts/mistralai_were-very-happy-to-announce-that-weve-entered-activity-7462415072930308097-rLdo",
        "date": "2026-05-19",
        "meta": "3,267 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "langchain-linkedin",
    "name": "LangChain",
    "url": "https://www.linkedin.com/company/langchain/",
    "image": "/images/feeds/linkedin/langchain-linkedin.png",
    "bestFor": "Production AI agents",
    "metric": "518.5K followers",
    "facts": [
      "518.5K followers",
      "Official LangChain account"
    ],
    "about": "LangChain's LinkedIn feed covers the practical lifecycle of building, testing, deploying, and monitoring agents across its product ecosystem.",
    "take": "The account is useful for a steady stream of concrete agent-engineering patterns and implementation examples. It is vendor-specific and high-volume, so distinguish durable architecture lessons from product promotion.",
    "highlights": [
      {
        "title": "LangChain Coding Agent Spend Control",
        "url": "https://www.linkedin.com/posts/langchain_how-langchain-made-coding-agent-spend-predictable-activity-7472309306248323072--d1z",
        "date": "2026-06-15",
        "meta": "57 likes"
      },
      {
        "title": "Box Agent Built on Deep Agents Architecture",
        "url": "https://www.linkedin.com/posts/langchain_box-agent-is-built-on-deep-agents-it-lets-activity-7471245848652263425-juae",
        "date": "2026-06-12",
        "meta": "63 likes"
      },
      {
        "title": "Recursive Language Model Workflows in Deep Agents",
        "url": "https://www.linkedin.com/posts/langchain_how-to-use-rlms-in-deep-agents-activity-7478115204162138112-kVNK",
        "date": "2026-07-01",
        "meta": "61 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "elevenlabs-linkedin",
    "name": "ElevenLabs",
    "url": "https://www.linkedin.com/company/elevenlabsio/",
    "image": "/images/feeds/linkedin/elevenlabs-linkedin.png",
    "bestFor": "Generative voice and media products",
    "metric": "316.8K followers",
    "facts": [
      "316.8K followers",
      "Official ElevenLabs account"
    ],
    "about": "ElevenLabs' LinkedIn feed tracks its rapidly expanding voice, audio, agent, and avatar products through releases and deployment stories.",
    "take": "The account is a useful first-party product feed in a fast-moving media category and often shows concrete use cases. It is less useful for comparing voice quality, rights, or pricing across providers.",
    "highlights": [
      {
        "title": "ElevenAgents Voice AI India Educational Platform",
        "url": "https://www.linkedin.com/posts/elevenlabsio_explore-elevenagents-voice-ai-that-sounds-activity-7472566326302171136-Fwc8",
        "date": "2026-06-16",
        "meta": "335 likes"
      },
      {
        "title": "The Odyssey Audiobook Michael Caine Voice Release",
        "url": "https://www.linkedin.com/posts/elevenlabsio_today-were-releasing-the-odyssey-an-audiobook-activity-7475184144902348800-x8dZ",
        "date": "2026-06-23",
        "meta": "194 likes"
      },
      {
        "title": "ElevenCreative Introduces Avatars Feature",
        "url": "https://www.linkedin.com/posts/elevenlabsio_elevencreative-all-in-one-ai-creative-platform-activity-7470891314990002176-BdAE",
        "date": "2026-06-11",
        "meta": "159 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "llamaindex-linkedin",
    "name": "LlamaIndex",
    "url": "https://www.linkedin.com/company/91154103/",
    "image": "/images/feeds/linkedin/llamaindex-linkedin.png",
    "bestFor": "Data and document systems for AI agents",
    "metric": "281.6K followers",
    "facts": [
      "281.6K followers",
      "Official LlamaIndex account"
    ],
    "about": "LlamaIndex uses LinkedIn to publish product releases, benchmark examples, integrations, and implementation patterns for data-backed AI applications.",
    "take": "The feed is useful for developers working on document and retrieval-heavy agent systems, especially when posts include a benchmark or integration detail. Company events and promotion add noise to the otherwise technical stream.",
    "highlights": [
      {
        "title": "The Agent Open AI Pickleball Tournament",
        "url": "https://www.linkedin.com/posts/llamaindex_the-agent-open-ais-pickleball-tournament-activity-7469763513281073155-LmIc",
        "date": "2026-06-08",
        "meta": "49 likes"
      },
      {
        "title": "LlamaParse Platform n8n Node Verified Community Release",
        "url": "https://www.linkedin.com/posts/llamaindex_the-n8n-node-for-the-llamaparse-platform-activity-7476304543543095297-bdFK",
        "date": "2026-06-26",
        "meta": "43 likes"
      },
      {
        "title": "Anthropic Fable 5 ParseBench Document Understanding Test",
        "url": "https://www.linkedin.com/posts/llamaindex_day-0-anthropic-fable-5-in-parsebench-we-activity-7470268027129724928-0AEJ",
        "date": "2026-06-10",
        "meta": "40 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "deepseek-ai-linkedin",
    "name": "DeepSeek AI",
    "url": "https://www.linkedin.com/company/deepseek-ai/",
    "image": "/images/feeds/linkedin/deepseek-ai-linkedin.png",
    "bestFor": "DeepSeek model releases",
    "metric": "191.4K followers",
    "facts": [
      "191.4K followers",
      "Official DeepSeek account"
    ],
    "about": "DeepSeek's LinkedIn presence is a sparse first-party stream for model releases and technical announcements from the Chinese AI lab.",
    "take": "Use the account to confirm official releases, then rely on technical reports, repositories, and independent evaluations to understand the models.",
    "highlights": [
      {
        "title": "DeepSeek-V3.2-Exp Model Release",
        "url": "https://www.linkedin.com/posts/deepseek-ai_introducing-deepseek-v32-exp-our-latest-activity-7401462255881748480-3Her",
        "date": "2025-12-02",
        "meta": "1,534 likes"
      },
      {
        "title": "DeepSeek-V3.2-Exp Model Release",
        "url": "https://www.linkedin.com/posts/deepseek-ai_introducing-deepseek-v32-exp-our-latest-activity-7378371456860262400-wwSo",
        "date": "2025-09-29",
        "meta": "1,284 likes"
      }
    ],
    "platforms": [

    ]
  },
  {
    "slug": "cognition-linkedin",
    "name": "Cognition",
    "url": "https://www.linkedin.com/company/cognition-ai-labs/",
    "image": "/images/feeds/linkedin/cognition-linkedin.png",
    "bestFor": "Devin and coding agents",
    "metric": "76.1K followers",
    "facts": [
      "76.1K followers",
      "Official Cognition account"
    ],
    "about": "Cognition's LinkedIn feed documents Devin product development, coding-agent evaluations, customer deployments, and changes in software-engineering work.",
    "take": "The account is a useful first-party source for new Devin capabilities and real deployment examples. Benchmark and productivity claims are central to the company narrative and need independent interpretation.",
    "highlights": [
      {
        "title": "FrontierCode: New Coding Evaluation Benchmark",
        "url": "https://www.linkedin.com/posts/cognition-ai-labs_introducing-frontiercode-a-new-coding-eval-activity-7469845270013480960-avzx",
        "date": "2026-06-08",
        "meta": "254 likes"
      },
      {
        "title": "Devin Fusion Hybrid Model for Agentic Coding",
        "url": "https://www.linkedin.com/posts/cognition-ai-labs_today-were-introducing-devin-fusion-a-new-activity-7477391045949800449-wWcn",
        "date": "2026-06-29",
        "meta": "187 likes"
      },
      {
        "title": "Devin AI Changes Engineer Workflows at AHEAD",
        "url": "https://www.linkedin.com/posts/cognition-ai-labs_devin-is-changing-how-engineers-work-at-ahead-activity-7475218891552768000-Af4h",
        "date": "2026-06-23",
        "meta": "112 likes"
      }
    ],
    "platforms": [

    ]
  }
];

export const otherFeeds = [
  {
    "slug": "sahar-mor-linkedin",
    "name": "Sahar Mor",
    "url": "https://www.linkedin.com/in/sahar-mor/",
    "image": null,
    "bestFor": "Curated AI research and builder tools"
  },
  {
    "slug": "scott-wu-linkedin",
    "name": "Scott Wu",
    "url": "https://www.linkedin.com/in/scott-wu-8b94ab96/",
    "image": "/images/feeds/linkedin/scott-wu-linkedin.png",
    "bestFor": "Building and measuring coding agents"
  },
  {
    "slug": "beautiful-ai-linkedin",
    "name": "Beautiful.ai",
    "url": "https://www.linkedin.com/company/beautiful-ai/",
    "image": "/images/feeds/linkedin/beautiful-ai-linkedin.png",
    "bestFor": "AI-assisted presentations"
  },
  {
    "slug": "wispr-flow-linkedin",
    "name": "Wispr Flow",
    "url": "https://www.linkedin.com/company/wisprflow/",
    "image": "/images/feeds/linkedin/wispr-flow-linkedin.png",
    "bestFor": "AI dictation product updates"
  },
  {
    "slug": "willow-voice-linkedin",
    "name": "Willow",
    "url": "https://www.linkedin.com/company/willowvoice/",
    "image": "/images/feeds/linkedin/willow-voice-linkedin.png",
    "bestFor": "AI voice input for professionals"
  },
  {
    "slug": "aqua-voice-linkedin",
    "name": "Aqua Voice",
    "url": "https://www.linkedin.com/company/aqua-voice/",
    "image": null,
    "bestFor": "AI dictation for professional work"
  }
];

export const BestForChip = ({ children }) => (
  <span className="inline-flex min-w-0 max-w-full items-center rounded bg-emerald-50 px-1.5 py-[2px] dark:bg-emerald-500/10">
    <span className="truncate text-xs font-medium leading-4 text-emerald-700 dark:text-emerald-400">{children}</span>
  </span>
);

export const Metric = ({ item }) => item.metric ? (
  <span className="inline-flex items-center gap-1 text-[13px] leading-5 uai-ink-muted">
    <span className="tabular-nums uai-ink-strong">{item.metric}</span>

  </span>
) : null;

export const FeedRow = ({ item }) => (
  <a href={item.url} target="_blank" rel="noreferrer" className="group flex items-center gap-3 px-4 py-3 no-underline transition-colors hover:bg-zinc-50/60 dark:hover:bg-white/[0.02]">
    {item.image && <img src={item.image} alt="" noZoom className="h-10 w-10 shrink-0 rounded-lg object-cover" />}
    <span className="min-w-0 flex-1">
      <span className="flex min-w-0 flex-wrap items-baseline gap-x-1.5">
        <span className="truncate text-sm font-semibold leading-5 uai-ink-strong">{item.name}</span>
        {item.metric && <span aria-hidden="true" className="uai-ink-muted">&#8729;</span>}
        <Metric item={item} />
      </span>
      <span className="block truncate text-[13px] leading-5 uai-ink-body">{item.bestFor}</span>
    </span>
    <span className="mr-1 flex shrink-0 items-center uai-ink-muted"><Icon icon="chevron-right" size={12} color="currentColor" /></span>
  </a>
);

export const FeedList = ({ items, visible = 7, toggleId }) => (
  <div className="not-prose mb-8 overflow-hidden rounded-xl border border-zinc-200 bg-white dark:border-white/10 dark:bg-white/[0.03]">
    <div className="divide-y divide-zinc-100 dark:divide-white/5">
      {items.slice(0, visible).map((item) => <FeedRow key={item.slug} item={item} />)}
    </div>
    {items.length > visible && (
      <span>
        <input type="checkbox" id={toggleId} className="fy-more-toggle peer sr-only" />
        <span className="hidden divide-y divide-zinc-100 border-t border-zinc-100 peer-checked:block dark:divide-white/5 dark:border-white/5">
          {items.slice(visible).map((item) => <FeedRow key={item.slug} item={item} />)}
        </span>
        <label htmlFor={toggleId} className="flex w-full cursor-pointer items-center justify-center border-t border-zinc-100 bg-zinc-50/60 px-4 py-2.5 text-[13px] font-medium text-zinc-500 transition-colors hover:text-zinc-800 dark:border-white/5 dark:bg-white/[0.02] dark:text-zinc-400 dark:hover:text-zinc-200">
          <span className="fy-more-open flex items-center gap-1.5">Show {items.length - visible} more accounts <Icon icon="chevron-down" size={14} /></span>
          <span className="fy-more-close items-center gap-1.5">Show less <Icon icon="chevron-up" size={14} /></span>
        </label>
      </span>
    )}
  </div>
);

export const FeedTabs = ({ children }) => (
  <fieldset className="fy-tabs">
    <legend className="sr-only">Feed category</legend>
    {children}
  </fieldset>
);

export const EntryCard = ({ item }) => (
  <section className="not-prose mb-5 rounded-xl border border-zinc-200 bg-white p-5 shadow-sm dark:border-white/10 dark:bg-white/[0.03]">
    <div className="flex items-center gap-3">
      {item.image && <img src={item.image} alt="" noZoom className="h-12 w-12 shrink-0 rounded-lg object-cover" />}
      <div className="min-w-0 flex-1">
        <div className="flex min-w-0 flex-wrap items-center gap-x-2 gap-y-0.5">
          <h2 className="m-0 max-w-full text-xl font-semibold leading-7 tracking-normal">
            <a href={item.url} target="_blank" rel="noreferrer" className="group inline-flex min-w-0 items-center gap-1.5 no-underline uai-ink-strong">
              <span className="min-w-0 break-words underline-offset-4 group-hover:underline">{item.name}</span>
              <span aria-hidden="true" className="flex shrink-0 text-zinc-400 transition-colors group-hover:text-zinc-900 dark:text-zinc-500 dark:group-hover:text-white"><Icon icon="arrow-up-right" size={12} color="currentColor" /></span>
            </a>
          </h2>
          <BestForChip>{item.bestFor}</BestForChip>
        </div>
        {item.facts.length > 0 && <div className="mt-1 flex flex-wrap items-center gap-x-2 text-[13px] leading-5 uai-ink-muted">
          {item.facts.map((fact, index) => <span key={fact} className="inline-flex items-center gap-x-2">{index > 0 && <span aria-hidden="true">&#8729;</span>}<span>{fact}</span></span>)}
        </div>}
        {item.platforms.length > 0 && <div className="mt-1 flex flex-wrap items-center gap-x-2 text-[13px] leading-5 uai-ink-muted"><span>Listen on</span>{item.platforms.map((platform) => <a key={platform.url} href={platform.url} target="_blank" rel="noreferrer" className="font-medium uai-ink-strong underline underline-offset-2">{platform.label}</a>)}</div>}
      </div>
    </div>
    <div className="mt-4 text-sm leading-[22px] uai-ink-body">
      <div>{item.about}</div>
      <div className="mt-3">{item.take}</div>
    </div>
    {item.highlights.length > 0 && <div className="mt-4">
      <div className="text-[13px] font-semibold uai-ink-strong">Popular posts</div>
      <div className="mt-1 divide-y divide-zinc-100 dark:divide-white/5">
        {item.highlights.map((highlight) => <a key={highlight.url} href={highlight.url} target="_blank" rel="noreferrer" className="group flex items-center gap-3 py-2.5 no-underline">
          <span className="shrink-0" style={{ paddingLeft: "2px" }}><Icon icon="linkedin" size={15} color="#0a66c2" /></span>
          <span className="min-w-0 flex-1 truncate text-sm font-medium leading-5 uai-ink-body">{highlight.title}</span>
          {highlight.meta && <span className="hidden shrink-0 text-xs tabular-nums uai-ink-muted sm:inline">{highlight.meta}</span>}
          <span className="shrink-0 uai-ink-muted"><Icon icon="arrow-up-right" size={11} color="currentColor" /></span>
        </a>)}
      </div>
    </div>}
  </section>
);

export const OtherFeedLine = ({ item }) => (
  <li className="flex items-baseline gap-2.5 text-sm leading-[22px] uai-ink-body">
    <span aria-hidden="true" className="relative -top-0.5 inline-block h-1.5 w-1.5 shrink-0 rounded-full bg-zinc-300 dark:bg-zinc-600" />
    <span>
      {item.image && <img src={item.image} alt="" noZoom className="relative -top-px mr-1 inline h-4 w-4 rounded-sm object-cover" />}
      <a href={item.url} target="_blank" rel="noreferrer" className="font-medium uai-ink-strong underline underline-offset-2">{item.name}</a>: {item.bestFor}.
    </span>
  </li>
);

<div className="not-prose -mt-5 mb-8 flex flex-wrap items-center gap-x-2 text-sm text-zinc-800 dark:text-zinc-200" style={{ paddingLeft: "2px" }}>
  <span className="inline-flex items-center gap-1.5"><Icon icon="clock-rotate-left" size={10} color="#27272a" /> Updated <time dateTime="2026-07-12">July 12, 2026</time></span>
</div>

LinkedIn is most useful for following AI when you want professional context around research, products, company strategy, and how teams are applying new tools at work.

We separate individuals from companies because they serve different purposes. Individual accounts add judgment and experience; company accounts are better for first-party product, research, and hiring updates.

## Best AI LinkedIn Accounts to Follow

<FeedTabs>
  <input type="radio" id="fy-tab-1" name="fy-tabs" defaultChecked />

  <input type="radio" id="fy-tab-2" name="fy-tabs" />

  <div className="fy-tablist not-prose">
    <label htmlFor="fy-tab-1">Individuals</label>
    <label htmlFor="fy-tab-2">Companies</label>
  </div>

  <div className="fy-panel" id="fy-panel-1">
    <FeedList items={individuals} visible={7} toggleId="fy-more-1" />

    {individuals.map((item) => <EntryCard key={item.slug} item={item} />)}
  </div>

  <div className="fy-panel" id="fy-panel-2">
    <FeedList items={companies} visible={7} toggleId="fy-more-2" />

    {companies.map((item) => <EntryCard key={item.slug} item={item} />)}
  </div>
</FeedTabs>

***

## How to choose

Choose a small mix rather than following every large account. Start with one researcher or educator, one operator or executive, and the companies whose products you actually use. Treat company posts as primary announcements, not independent analysis.

***

## Other LinkedIn Accounts to Consider

<ul className="not-prose m-0 my-4 flex list-none flex-col gap-2 p-0">
  {otherFeeds.map((item) => <OtherFeedLine key={item.slug} item={item} />)}
</ul>
