Research
Heilongjiang rice platform reports shorter breeding cycles with AI
A People’s Daily field report describes AI-assisted rice breeding in Jiamusi; the report does not provide test details for the claimed time savings.
Research.
Ideas, evidence and the questions shaping what comes next.
27 storiesPage 3 of 5
Research
Industry research estimates China’s mobile-robot market at RMB14.2 billion
A new Zhiyan Consulting industry article estimates China’s intelligent mobile-robot market for 2025 and describes a fragmented global vendor landscape.
Research
Fermi Universe Raises Seed Round for Quantum-Inspired AI Models on Classical Hardware
Chinese startup Fermi Universe has raised 100 million yuan in seed funding at a roughly 1 billion yuan valuation to develop quantum-inspired enhancements for large language models that run entirely on classical GPU infrastructure, according to a QbitAI interview originally published on September 27, 2026.
Research
Shanghai Startup Unitary Quantum Launches Desktop Hardware and AI-Agent Platform for Hybrid Quantum Computing
Unitary Quantum, a Shanghai Jiao Tong University spinout, unveiled UnitarySpark desktop hardware and the UnitaryLab 2.5 public beta at a September 20 event in Shanghai. The system pairs local heterogeneous computing with an AI-agent interface that translates natural-language scientific questions into quantum workflows, keeping all data on-premises.
Research
Alibaba DAMO Academy Publishes AI Model for Esophageal Cancer Screening on Non-Contrast CT in Nature Medicine
Alibaba's DAMO Academy and partner hospitals published a Nature Medicine paper in September 2026 describing EAGLE, an AI model that screens for esophageal cancer and precancerous lesions using routine non-contrast chest CT. Validated across more than 80,000 patients in three countries, the model achieved 90% sensitivity for cancer detection in external testing. Researchers emphasize the tool is intended to triage high-risk patients for confirmatory endoscopy, not replace it.
Research
Xiaomi's HySparse2 targets lower computing demands for long-context AI agents
Xiaomi researchers describe an attention design that reduces the computation and memory needed to process long inputs. Their paper reports gains over two research baselines, with important limits on what the results establish about real-world performance.
