Baichuan's Chuan Wang: In the Era of General AI, Building Models for Life, Creating Doctors for Humanity | Xingxin PORTFOLIO

Chuan Wang, CEO of Baichuan, shared a reflection on entrepreneurship today:

Two years since its founding, Baichuan has evolved from a frontier large language model company into a bold entrant in life sciences. Under Wang's leadership, the team has held fast to its founding purpose: "building models for life, creating doctors for humanity" — with the ultimate goal of deploying AI to crack medicine's hardest problems.

AI healthcare remains uncharted territory, with academia and industry still feeling their way forward. Baichuan has already pushed into the furthest reaches of this unexplored terrain. Understanding the mysteries of life is like scaling Everest, yet every breakthrough brings technology and life closer together.

To date, Baichuan has established connections with seven of Beijing's national medical centers. The long march of AI healthcare has begun, and Baichuan is on the path.

The following is Chuan Wang's personal reflection:

Today is April 10. Two years ago, convinced that breakthroughs in language AI signaled the arrival of artificial general intelligence, I founded Baichuan — becoming one of the first Chinese companies dedicated to AGI research. I am grateful to every colleague who has joined us along the way, bearing witness to and advancing this extraordinary era.

Over these two years, AGI has surged forward with the force of a storm, one wave cresting after another, each more gripping than the last. Whenever the public has questioned whether intelligence has hit a ceiling, a larger breakthrough has emerged. For the AI industry, for startups, for Baichuan, and for each of us as individuals, what is needed now is not merely technical foresight, but steadfastness in direction and unwavering belief in the path.

Baichuan has accomplished much in these two years.

First, our predictions on technological breakthroughs and industry direction have been borne out. On the technology front, I have long held that language is the central axis of intelligence, and that advances in language AI herald the era of strong AI. I also identified reinforcement learning-driven "slow thinking" as the second paradigm shift. Both have materialized. I further predict that "Coding" will bring the third and final paradigm shift. On the industry front, I have maintained that healthcare is among the greatest applications of the intelligence age, proposing that "doctors are the central axis of medicine" and putting this into practice by building AI doctors.

In the first two years, skeptics in the industry argued that medical AI was still far off. By 2025, investment fervor across the Pacific has already surged toward medical AI. Just last month, Bill Gates predicted AI would replace doctors, while Bloomberg forecast that Apple will launch an AI health coach to substitute for human physicians by 2026.

Second, our R&D and commercial deployment have shown solid results. The open-source Baichuan 1 and Baichuan 2, released in 2023, gained wide adoption — by some estimates, half of all domestically registered models that year were built on Baichuan's foundation. In 2024, we laid out AI general practice, AI pediatrics, and digital biomarkers; established capital ties with Luca and Xiao'er Fang; partnered with Tsinghua University's China Institute for Healthy Development to jointly research evaluation frameworks and service standards for intelligent general practitioners; and co-founded a Beijing municipal key laboratory with Beijing Children's Hospital. In early 2025, we released Baichuan-M1, a medically enhanced reasoning model; piloted our AI general practitioner in Haidian District; jointly launched the "Futang · Baichuan Pediatric Large Model" with Beijing Children's Hospital; and deployed our AI pediatrician to official duty at Beijing Children's Hospital.

Our work has also fallen short in several respects.

First, two years of relentless sprinting stretched our frontlines too thin, lacking sufficient focus. From general foundation models to medically enhanced reasoning models, from Baixiaoying and AI doctor applications to premature commercialization, we greatly increased organizational complexity.

Second, although our mid-2024 strategy meeting had clarified our healthcare focus, we did not sufficiently communicate our resolve and the required path in healthcare. We failed to push every team to deeply think through the "why" and "how" of creating medical value. As a result, some teams wavered and deviated in their objectives.

On this second anniversary, I want to explain our vision and business direction to every Baichuan colleague.

On April 10, 2023, in Baichuan's founding announcement, I wrote: "I have embarked on a new expedition: dedicating the next twenty years to contributing to the advancement of life sciences and medicine, to making a difference for public health. The core path is constructing mathematical models for life and health, and action has already begun. The emergence of ChatGPT will also become an important catalyst for this new expedition."

Thus, building models for life, creating doctors for humanity — this is Baichuan's mission. The "Bai" in Baichuan is a homophone for "bio" — this aspiration was set from day one.

The bottleneck in medicine's development is real-world data across the full course of disease and the full life cycle. The bottleneck in healthcare delivery is the training and supply of quality doctors. In the AGI era, we will break through on both fronts. Large model intelligent agents can provide unlimited AI doctor supply, collaborating with human doctors in a "dual-doctor model" — giving every doctor an AI assistant, every patient a dedicated AI doctor that accompanies them long-term and records their individual data.

To realize this vision, our concrete path is "create doctors — transform pathways — advance medicine." Create doctors: focus on the most broadly needed and most scarce specialties, general practice and pediatrics. Transform pathways: deploy unlimited AI doctor supply to strengthen grassroots care and tiered diagnosis and treatment, aligned with national policy; develop digital biomarkers to extend care scenarios into the home, achieving hospital-at-home and digital companionship. Advance medicine: the data generated through these services will fuel breakthroughs in clinical medicine and epidemiology, converging with basic medicine to establish life models, ultimately progressing toward individualized precision medicine.

Healthcare is unlike the internet's demand-driven, product-research coordination model. Fortunately, by the end of 2024, we had assembled a high-caliber medical department capable of guiding and supporting Baichuan's medical model R&D and driving partnerships with leading medical institutions. Of Beijing's thirteen national medical centers, seven have already established connections with Baichuan.

Building on this foundation, we are developing applications and services across four areas.

Baixiaoying: A medically enhanced large model, infused with medical reasoning and evidence-based medicine, providing authentic, authoritative responses with citations — achieving "multidisciplinary team in one," becoming a powerful assistant for hospitals and doctors in clinical practice and research.

AI Pediatrics: Co-developed with Beijing Children's Hospital, a high-level pediatric large model to serve doctors; addressing the severe shortage of pediatricians by "creating 1 million pediatric doctors" to empower grassroots care; and directly providing health guidance to parents and children.

AI General Practice: Strengthening grassroots care and tiered diagnosis and treatment, becoming the AI family doctor for the nation — under precision medicine, enabling early screening, early diagnosis, and health management.

Precision Medicine: Building on Luca's accumulated capabilities and breakthroughs, developing digital biomarkers and digital companion services for different diseases, enhancing the diagnostic capabilities of AI Pediatrics and AI General Practice, and forming solutions with pharmaceutical and insurance partners.

In the AGI era, both technological progress and application directions are filled with non-consensus. AI healthcare is especially virgin territory, with academia and industry still exploring. We have already pushed into the furthest reaches of this uncharted frontier. Turning ideals into reality, transforming uncertainty into certainty — this is true entrepreneurship. The industry has undergone profound change in two years, especially with DeepSeek's open source (ps. salute to DeepSeek!). Yet Baichuan's mission remains unchanged, and more achievable than ever. At this two-year mark, we must reaffirm our founding purpose, reorganize for entrepreneurship, and reshape our entrepreneurial culture.

Going forward, we need to focus and concentrate around the directions above, eliminating extraneous moves, thinking deeply, staying firm without wavering. We need to reduce organizational complexity, become flatter with fewer layers, making information flow and decision-making at Baichuan more fluid. We need the nourishment of idealism — believing in Baichuan's ideals, believing in the meaning of solving problems and creating value.

The AGI era is a great era, greater than the internet age. Not only is human civilization ascending to new heights, but every individual is becoming stronger through AI's empowerment, better able to find "who I am." Building models for life, creating doctors for humanity — this is a mission to be proud of. I am grateful to every colleague participating in this journey, to industry partners and shareholders who support and help Baichuan. I believe this endeavor can bring us both spiritual and material abundance, and leave its mark on human history.

Chuan Wang, CEO of Baichuan April 10, 2025

Heart Capital was founded in 2022 as an early-stage Chinese venture capital fund focused on technology and digital transformation. The team is led by Yan Han, founding partner of Lightspeed China, alongside core investors, a CFO, and seasoned investors from industry. The team's past investments include Series A stakes in Xpeng Motors (NYSE: XPEV, 09868.HK), Full Truck Alliance (NYSE: YMM), as well as FinVolution (NYSE: FINV), RoboSense (02498.HK), Baichuan, Manman Lengyun, Fan Deng Reading, World Logistics, Weina Xingkong, LandSpace, Lanhu, Starfield, and others. Rooted in China with a global outlook, Heart Capital seeks genuine value in non-consensus. The firm honors the value of people and champions the potential of the heart, aspiring to accompany more young Chinese entrepreneurs in strengthening China and reaching the world.