As it has been difficult to train a large number of people to be well versed in both brain science and information technology for developing brain-inspired AGI, WBAI has initiated BRA-driven development, a brain-inspired software development method, in which BRAs are created as design specifications with accumulated neuroscientific findings, and software will be developed and evaluated based on the BRAs. In this approach, the immediate technical objective of WBAI, which promotes the development of AGI referring to the entire brain, will be the construction of BRA data.
Last year, the BRA data format was standardized, and the formulation of BRA-based development made significant progress. This year, we shall disseminate the method and encourage researchers interested in computational cognitive neuroscience to use it.
This year, as before, we continue our education and research & development businesses in the following activities.
Education Business
The objective of the education business is to develop and increase the number of interdisciplinary personnel with different expertise in AI/ML, neuroscience, and cognitive science necessary for R&D in the WBA approach over the long-term. We continue our regular activities by holding seminars and an annual symposium to raise public interest.
We hold a hackathon (online international competition) to increase the number of researchers, engineers, and students who are interested in the WBA. We shall also publicize the progress of the WBA by participating in and cooperating with external academic events, and by cooperating and exchanging information with external academic organizations. We shall also cooperate with external researchers, engineers and groups to develop human resources and promote their involvement.
We shall call for the WBAI Incentive Award to recognize those who have made significant achievements in promoting the development of brain-inspired AGI technology, thereby revitalizing the community.
Research and Development Business
The objective of the R&D business is to accelerate democratic AGI research by promoting external research in the WBA approach: we shall conduct our activities in a way that does not compete with existing WBA research institutes.
In FY2021, we continue to accumulate neuroscientific knowledge in the BRA format, which consists of BIF, the description of the anatomical structure of the brain, and HCD, the description of hypothetical functional modules consistent with the BIF. BRAs will be used to facilitate the creation of brain-inspired AI for specific tasks. We will also develop techniques to evaluate the neuroscientific validity of brain-inspired AGI while using BRAs to evaluate the implementation of brain-inspired AI.
We also continue to work on integrated software platforms that will be the foundation for the construction of WBA, and on activities to make the brain-inspired AGI that we are promoting more beneficial.
Most of these research and development activities will be funded by external research organizations.
Budget for the fiscal year
The planned income is approximately 2.11 million yen, which includes 1.76 million yen in membership fee income as well as fees from seminars. As the membership fees from the supporting members were 3.23 million yen in FY 2020, the planned income from them has decreased significantly; this is because the three founding supporting members who have supported us have reached their five-year term. The total of the planned income for the current fiscal year and the 9.16 million yen carried over from the previous fiscal year is approximately 11.27 million yen.
Expenditures will be approximately 3.05 million yen in total, including approximately 0.83 million yen for administrative expenses and approximately 2.22 million yen for operating expenses, including event expenses and R&D expenses (in relation to the planned income for the current fiscal year, there will be a deficit of 0.94 million yen). In the FY2020 budget, we planned to spend about 3.73 million yen in total, including about 0.97 million yen for administrative expenses and about 2.77 million yen for operating expenses.
Milestones
Brain-inspired AGI development based on the whole-brain architecture approach has been embodied in the form of BRA-driven development, in which BRA as the design specification is created with the constraint of brain scientific findings, and software is developed and evaluated by referring to BRA. In order to complete brain-inspired AGI with this methodology, the following five milestones would have to be achieved:
- Completion of BIFs (brain information flow diagrams) covering all the brain regions: to integrate multiple mammalian neuroscientific findings to build a chimeric BIF that covers (almost) the entire brain.
- Completion of HCDs (hypothetical component diagrams) for typical functions: to construct at least one HCD for every BIF.
- Initial implementation of the entire WBA: to create software in which almost all regions of the brain are implemented and integrated based on the BRA.
- Automated architecture exploration: to run the brain-inspired AI software in a virtual environment to automate testing and enable the efficient comparison and examination of various architectures.
- Coverage expansion of HCDs: to cover typical cognitive tasks that can be solved by humans and animals.
- Completion of the WBA system: to create software which can solve typical human tasks in the same way as the brain, consisting of almost all the components corresponding to the neural circuits of the brain.
With these milestones, we will focus on building BRA data for the entire brain (the first and second items) as well as evaluating the feasibility of the third item (implementation) for the time being. At this stage, we will be able to accelerate R&D by involving researchers interested in computational cognitive neuroscience. Thus, we would first like to promote BRA-based software development through academic presentation and human resource development. By steadily continuing these activities, we will be able to build a technological foundation for the development of brain-inspired AGI in the near future, and to set about the second half of the milestones. Once the completion of brain-inspired AGI is in sight, safety considerations will become more important; we will continue discussions on how to realize the brain-inspired AGI that will eventually be reached in a beneficial and safe manner.
We would like to continue our efforts based on the perspective above, and we appreciate your understanding and support for our endeavor.
Japanese