Ling Fan范凌

Past research · A quantitative study and a qualitative study

Brain–Machine Ratio (BMR)

In a creative process shared by people and machines, which part should each take, and how can that division be measured?

The Brain–Machine Ratio (BMR) asks how creative work should be divided between the designer's judgment and the machine's capacity.

It is studied in two ways, side by side. The quantitative study measures how the work divides today, task by task. The qualitative study traces how that division came to be, through six centuries of creative tools.

Paper (IEEE MIPR 2021) ↗

Two studies

BRAIN · WANTS TOBRAIN · HAS TOMACHINE · UNABLEMACHINE · ABLEWhat humans are good atWhat people want to doWhat people would rather not doWhat machines are good at2 · THE BRAIN’S STRENGTHSCreative ideationCommunication1 · HUMAN EXPERIENCEMaterial gathering3 · MACHINE POTENTIALManagementNon-repetitive physical work4 · MACHINE OPTIMIZATIONInformation processingRepetitive physical workH.I. · the human brainA.I. · the machineTOP HALFWhat people should leadBOTTOM HALFWhat to hand to machinesLEFT HALFHigh brain–machine ratioRIGHT HALFLow brain–machine ratio

(a)

Quantitative study

The BMR quadrant sorts creative tasks by two questions: do people want to do the task or only have to, and can machines do it? The upper half is what people should lead; the lower half is what they can hand to machines. BMR 1.0 measures capability, the ratio of human to machine input in each task. BMR 2.0 multiplies that ratio by subjectivity, the will people bring to a task. BMR 3.0 is about trust: AI that automates what is easy and safe to delegate, and augments what is hard and cannot be delegated. Design teams at Alibaba and Tencent have used it to build teams and organize creative workflows (IEEE MIPR 2021).

WDCC 2026 Theme Exhibition, “Designing Generation: From AI-Driven Design to Designing AI”, Shanghai.

(b)

Qualitative study: A History of Creative Tools

From perspective and photography to computers, software and AI, each change in tools has altered the threshold, the process and the possibilities of creative work. The study follows three threads from 1400 to 2026: technology brings new capabilities, tools turn them into practice, and ideas keep asking what creating means. The first edition was completed in 2023; the second, in 2026, adds generation, reasoning, agents and world models. Its nine-panel timeline was shown at the WDCC 2026 Theme Exhibition in Shanghai, where Ling served as Co-Chief Curator. Funded by the National Social Science Fund of China (2024–2027).

(a) Quantitative study · the BMR quadrant

BRAIN · WANTS TOBRAIN · HAS TOMACHINE · UNABLEMACHINE · ABLEWhat humans are good atWhat people want to doWhat people would rather not doWhat machines are good at2 · THE BRAIN’S STRENGTHSCreative ideationCommunication1 · HUMAN EXPERIENCEMaterial gathering3 · MACHINE POTENTIALManagementNon-repetitive physical work4 · MACHINE OPTIMIZATIONInformation processingRepetitive physical workH.I. · the human brainA.I. · the machineTOP HALFWhat people should leadBOTTOM HALFWhat to hand to machinesLEFT HALFHigh brain–machine ratioRIGHT HALFLow brain–machine ratio
BMR 1.0 · CAPABILITYBMR 1.0 =Human inputMachine inputManagement9%Creative ideation18%Communication20%Non-repetitive physical work25%Material gathering64%Information processing69%Repetitive physical work78%Fill and %: share machines can takeCircle size: share of the designer’s timeBMR 2.0 · SUBJECTIVITYBMR 2.0 = BMR 1.0 × subjectivitysubjective willBrain: time people investMachine: how far it can be automatedBMR 3.0 · TRUSTBMR 3.0 = human-centered AINot delegable to machinesDelegable to machinesEasy for humansHard forhumansAugmentationAutomation

Redrawn from Ling Fan, From Universality of Computation to the Universality of Imagination: A Catalog on Design and Artificial Intelligence (Tongji University Press, 2019).

(b) Qualitative study · exhibition and timeline

The timeline of creative tools on the exhibition walls, WDCC 2026.
Timeline panel 1Timeline panel 2Timeline panel 3Timeline panel 4Timeline panel 5Timeline panel 6Timeline panel 7Timeline panel 8Timeline panel 9
The nine timeline panels, 1400 to 2026, each in three threads: technology, tools and ideas (second edition, 2026; in Chinese).

Facts

First set out in
From Universality of Computation to the Universality of Imagination (2019)
Paper
The Brain-Machine-Ratio model for designer and AI collaboration (IEEE MIPR 2021)
Used by
Design teams at Alibaba and Tencent
Timeline
nine panels, 1400 to 2026; first edition 2023, second edition 2026
Exhibited
WDCC 2026 Theme Exhibition, "Designing Generation: From AI-Driven Design to Designing AI", World Design Cities Conference, Shanghai; Ling was Co-Chief Curator of WDCC 2026
Funding
National Social Science Fund of China, Research on the Evolution of Design Tools in the Era of Artificial Intelligence (2024–2027); Ministry of Education (2020); Shanghai art and technology program (2019)
Timeline team
Ling Fan with Li Dan, He Ziming, Wu Pengfei and Zhong Siyuan (2023); Ling Fan with Xia Lei and Zhou Zhiyuan (2026)
Research line
Human–AI collaboration, Design AI Lab

Next: Work 01 · Subjective World Model →