From August 21 to 22, the 2026 Science Intelligence Congress (AI for Science Congress 2026) was held at the Zhongguancun International Innovation Center in Beijing.
As part of the Zhongguancun Forum series, this congress, themed "AI Transforming Scientific Research Paradigms," brought together nearly 50 academicians and featured over 100 high-level reports showcasing significant achievements in the field of scientific intelligence. It stands as the highest-level academic gathering in the domestic AI4S field, sending a powerful signal of the times.
This is a shift in the "center of the AI era."
On this new paradigm map, OxTium Technology is deeply engaged as a representative frontier AI4S enterprise: OxTium co-founder & COO Zhou Tianyao organized and planned the selection of all exhibiting enterprises, recruitment of investment institutions, and exhibition area layout for this congress, while founder & CEO Jin Yongcheng delivered a keynote speech titled "AI for BioScience – Physical Intelligence Infrastructure for the Bio-AGI Era." This time, OxTium was also invited to set up a core exhibition area, showcasing cutting-edge technologies and industrial achievements, resonating with the AI4S era.
Signal of the Times:
Scientific Intelligence Becomes the New AI "National Narrative"
This congress is a highlight of the 2026 Zhongguancun Forum series, themed "AI Transforming Scientific Research Paradigms," co-hosted by the Zhongguancun Science City Management Committee, the Beijing Academy of Artificial Intelligence (AISI), and the China Science and Technology Development Foundation.
The congress's high profile is unparalleled in the domestic AI4S field: nearly 50 academicians from the Chinese Academy of Sciences and the Chinese Academy of Engineering, over 80 young scholars, and more than 2,500 representatives from government, industry, academia, and research deeply participated. Over 100 high-level reports showcased a batch of important achievements, with exchanges centered on the three-tier logic of "infrastructure – paradigm innovation – industrial application." The plenary session was co-chaired by academicians E Weinan and Bao Weimin, with academicians Wang Jian, Tang Chao, Sun Ninghui, and Xie Xiaoliang serving as co-chairs.
This congress conveyed a clear signal: the center of AI discussions is rapidly shifting from language models and generative applications to scientific intelligence.
Academician E Weinan said in his opening remarks, "The Science Intelligence Congress has been held for several years, but it was previously organized by non-governmental entities; this is the first time it is officially organized." This marks the formal elevation of AI for Science from an academic methodological movement to a national narrative of building a "globally influential scientific intelligence innovation hub."
In his report, Academician E Weinan pointed out that AI for Science has entered the "second half," with the focus shifting from methodological exploration to systemic transformation of the scientific research system.
Co-organizing with MiraclePlus
AI4S Project Roadshow
At this event, OxTium co-founder & COO Zhou Tianyao was at the forefront of the congress organization.
Zhou Tianyao, together with MiraclePlus, co-organized and planned the AI4S project roadshow for this congress. Centered on the core proposition of "how scientific intelligence moves from the lab to industry," they selected and organized a group of frontier AI4S projects to take the stage, building a bridge between research achievements and industrial capital.
"The core of AI for Science is not to replace scientists with AI, but to provide scientists with a new paradigm for scientific discovery. When large models can understand experimental data, drive hypothesis generation, and participate in the entire research process, the speed of scientific discovery will undergo a qualitative change," Zhou Tianyao said in his hosting remarks at the roadshow.
As Zhou Tianyao said, OxTium Technology's presence at the center of the AI4S event is not about following trends, but because what we started three years ago has just been caught up by the times. OxTium is not only a frontier technology explorer but also a promoter of paradigm change.
Building the Physical Intelligence Infrastructure
for the Bio-AGI Era
OxTium Technology founder & CEO Jin Yongcheng delivered a speech titled "AI for BioScience – Physical Intelligence Infrastructure for the Bio-AGI Era," systematically elaborating OxTium's technical roadmap.
At the cognitive foundation level, OxTium has self-developed the GeneLLM series of large models – the world's first multi-omics large model pre-trained on raw data, providing generalization capabilities for biopharmaceuticals, healthcare, and biomanufacturing.
At the engineering implementation level, the BioFord Harness multi-agent engineering adopts a three-tier architecture of decision brain, execution engine, and data recycling, coordinating all laboratory resources, interfacing with mainstream equipment from Tecan, Thermo Fisher, Illumina, and MGI, translating DSL decisions into real experiments, and achieving 7×24h uninterrupted execution.
At the evolutionary closed-loop level, the BioTide physical world model evolution framework starts with human demonstrations, validates during robot execution, and continuously optimizes through multi-round feedback, forming proprietary evolutionary assets; Data Engineering integrates dry and wet data to build a closed-loop data flywheel.
Jin Yongcheng further introduced OxTium's commercialization progress on both B2B and B2C fronts: for research institutions and enterprises, autonomous laboratory products have achieved large-scale deployment; for researchers, multiple products have completed commercial validation, with more new products accelerating.
In conclusion, he stated: "OxTium Technology aims to build the physical intelligence infrastructure for the Bio-AGI era, liberating scientists from repetitive labor and autonomously expanding the boundaries of human understanding of the physical world."
The Next Decade of Scientific Intelligence
OxTium's "AI for Bio Loop" Attracts Attention
OxTium's booth attracted attention and drew visitors from all sectors.
Lin Runhua, Secretary-General of the "Academic Science Association" Scientific Intelligence Expert Committee and President of Science and Technology Review, visited the OxTium booth, engaging in cordial conversations with the team, learning in detail about the BioScience large models and autonomous laboratory solutions, and expressing interest in OxTium's "AI for Bio Loop" closed-loop path.
Figure: Lin Runhua, Secretary-General of the "Academic Science Association" Scientific Intelligence Expert Committee and President of Science and Technology Review, and his delegation visit the OxTium booth
Visitors at the booth were in a steady stream, with attendees from universities, research institutes, and industry stopping to learn about OxTium's closed-loop solutions from cognition to execution to evolution, engaging in in-depth exchanges and discussions with the team, and multiple institutions expressed intentions for cooperation on-site.
OxTium embeds an "enterprise perspective" into the decision-making flow of the academic congress – a microcosm of what Academician E Weinan called for: "connecting talent, innovation, market, and capital from an industrial perspective."
As AI moves from the frenzy of language models to the deep waters of scientific discovery, from "computing" to "doing," a new era centered on scientific intelligence is beginning. If this congress is its opening ceremony, OxTium Technology has already secured a place on the stage.
In the next decade, the speed of scientific discovery will be redefined.
OxTium Technology will continue to deepen its work in BioScience large models and autonomous laboratory infrastructure, joining hands with academia and industry to embrace the Bio-AGI era.