Linear Capital's 2026 AGM Concludes Successfully in Shanghai

On November 9–10, Linear Capital held its 2026 Annual General Meeting (Linear Capital 2026 AGM) on the Shanghai Bund. Over 200 top global LPs and founders from Linear's portfolio gathered to reflect on and look ahead at the opportunities and challenges reshaped by frontier technologies, with AI at the forefront.

At this gathering, we saw people earnestly working on answers: enabling robots to complete one more reliable task, getting new materials validated faster, giving ordinary users more freedom to go about their lives, and making technological capabilities accessible to more people.

The future still holds no standard answers, but Linear hopes to work alongside founders to expand the boundaries of intelligence while giving more people access to capability, opportunity, and dignity — turning optimism about the future into concrete actions worth taking today.

At last year's AGM, we asked, "To be or not to be, in the AI-dominated future." A year later, AI has advanced considerably further. This year, we pushed the question deeper: "The AI Heaven or The AI Hell" — as AI grows ever more powerful, what future awaits humanity?

Facing this question, our emotions may exist in a quantum superposition of fear and excitement.

This was also a genuine feeling at this year's conference. We saw exciting progress in models, robotics, and scientific computing, while also hearing founders discuss the concrete difficulties behind the technology: whether a dexterous hand can withstand continuous use, whether a family would feel comfortable entrusting tasks to a robot, whether an experimental result can truly make it out of the lab. These questions are determining how technology meets everyday life.

The day's team presentations began with Linear's own year in review. Over 12 years, Linear has accompanied nearly 200 companies in their growth. While maintaining a team of under 20 people, Linear has adopted a nearly radical approach to making its own workflow sufficiently AI-native, keeping pace with advancing technology and portfolio development.

Internally, we've gradually refined a comprehensive AI system through practice, deeply integrating AI into pre- and post-investment work so that more energy can be devoted to engaging with founders and making judgments that require human insight. This year, through video podcast content like "Between the Lines" and other brand initiatives, we also hope to bring the authentic experiences and reflections from the entrepreneurial journey to young founders we haven't yet met.

Looking ahead, Linear's investment team shared observations and investment theses around three directions — Agentic AI hardware and software, AI Infrastructure, and AI for Science — while also casting their gaze toward earlier-stage technological frontiers including brain-computer interfaces, space exploration, and quantum computing. A thread running through our thinking: What are the most worthwhile problems to solve today, and what kind of people can actually solve them?

In the AI Infrastructure direction, we discussed the growth in intelligent demand and the practical bottlenecks supporting it. As AI takes on more tasks, inference efficiency, storage, and hardware-software coordination become increasingly critical. Linear seeks teams that understand model evolution and can turn technology into products, building globally competitive capabilities in key links.

In the Physical AI direction, from new robotic forms to critical components, Linear hopes to find teams that can unlock new applications and solve industrial problems, as well as technologies and capabilities in the supply chain that are difficult to replace.

In the Agentic AI direction, we care whether AI can achieve "closed-loop task completion." Who a product serves, what counts as done, and how it improves through real tasks and feedback — these all need to be thought through from the start of a venture. Software needs to bridge the gap between capability and daily use; hardware has the opportunity to more naturally understand the user's environment. Teams are valued for their technology, but also for their founders' understanding of people: what's worth doing, what experience will make users want to return again and again. And design is precisely what enables technology to enter people's lives and be used continuously.

In the AI4S direction, we discussed how scientific discovery is changing. When generating hypotheses becomes easier, reliable experimentation and validation become more precious. From algorithms and models to sustainable research loops, outstanding teams need to simultaneously understand science, AI, and engineering — knowing how to make one experiment help the next one go better. For more frontier exploration, Linear also insists on grounding long-term vision in the next verifiable milestone: what risks are addressed, how much time and investment is needed, and how much closer to customers and real value.

In the afternoon, portfolio companies took the floor with their own practices. In the Embodied Intelligence Commercialization forum, Harry Wang, Founder and CEO of Linear Capital, joined Zhao Yuekai, Co-founder of Agile Robots, Dalong Du, Founder of Zhangyu Power, and Yufeng Zhang, Founder of Anyverse Dynamics, in discussion around real-world scenarios, scaled delivery, and globalization. Moving from technology demonstrations to sustained value creation requires not just breakthrough capabilities, but also the patience to refine products repeatedly at customer sites.

In the New Embodied Forms forum, Cong Wang, Founder of D-Robotics, Wu Shulin, Founder of Zhengqi Future, and Wei Chen, Founder of Xieyue Intelligence, brought the discussion into daily life: if a robot appeared in your home tomorrow, what would you want it to do, and why would you let it stay? From mobility to household chores, the speakers shared their thinking on understanding users and making trade-offs on functionality. Safety, trust, and user experience give "robots entering the home" more concrete meaning.

In the AI Infrastructure forum, industry veteran Bill and Xu Ziyang, Co-founder of Xingan Future, discussed new demands in inference and storage, hardware-software co-design, and how AI is changing chip development approaches. Rapid technological evolution also poses new requirements for entrepreneurial teams: understanding the changes while identifying fundamental problems worth long-term investment.

In the company presentation sessions, teams from Anyverse Dynamics, Deep Principle, and Yuanqi Quantum shared their explorations in embodied intelligence going global, AI-driven materials R&D, and quantum computing respectively.

Ideas that were just getting started a year ago now have new products, experiments, and customer feedback; more early-stage technologies are also finding their first paths to industry. These incremental advances show us how long-term visions are becoming reality bit by bit.

At the AGM's close, Harry delivered a keynote on "The AI Heaven or The AI Hell," pushing the discussion beyond technology and business toward questions that merit deeper reflection: As intelligence becomes ever more abundant, where do we want it to take the world? AI expands the boundaries of human capability, while also making unavoidable the questions of who defines goals, how value is distributed, and how human judgment and choice are preserved. Whether technological progress translates into human welfare depends on the choices behind every product and decision made today.

This also gives early-stage investment longer-term meaning. Supporting what kind of founders, encouraging what kind of products — this concerns not just commercial returns, but also participates in shaping how technology enters society. Linear hopes to work alongside founders to expand the boundaries of intelligence while giving more people access to capability, opportunity, and dignity, turning optimism about the future into concrete actions worth taking today.

In the face of constantly shifting markets, Linear's stance is "cautiously optimistic, disciplinedly active." From an idea not yet fully understood, to product, to company, to genuine value realization — the path requires openness, action, patience, and also fact-respecting judgment and discipline.

Returning to this year's AGM theme, does AI's future look more like Heaven or Hell? At the end of Harry's keynote, after hearing those reflections on capability, risk, and the human condition, more people in the audience ultimately chose that we're in "AI Heaven."

The future still holds no standard answers. But at this gathering, we saw people earnestly working on answers: enabling robots to complete one more reliable task, getting new materials validated faster, giving ordinary users more freedom to go about their lives, and making technological capabilities accessible to more people.

During the portfolio company visit day, attendees toured SynMeta, Tashi Zhihang, and RIVR, gaining close-up understanding of portfolio companies' technological progress, product development, and industrialization practices. These concrete advances are our reason for optimism, and also what Linear is willing to continue investing in alongside founders.

More highlights from Linear Capital's 2026 AGM will be shared gradually through our WeChat official account and other channels — stay tuned.