What will researchers get from IIAI if their colleges participate in IIAI?
Draft text summarizing IIAI's vision of services provided to UI researchers.
- AI/Machine Learning environment
- Access to a well-designed and well-maintained deep learning/GPU/AI environment
- Engineering Computing Services in charge with a staff person being hired to fulfill this role
- Environment will
- include well-tested instructions for users
- only allow access to users who passed entry-level user tests
- operate based on a well-developed use policy
- Environment will allow
- Development of applications and initial testing in IIAI development labs
- Access to staging storage space for AI development, well organized with group access for research teams
- Relatively seamless transition to run production-level jobs on the Argon HPC
- Access to state-of-the-art, cutting-edge environment
- Environment will be kept at the forefront of technology
- Computational upgrades will be financed from these resources
- Scalable environment of sufficient capacity
- As the use need will grow, we will have resources to respond to this growth quickly and operatively, also in a location-specific manner
- AI/GPU-enabled Argon HPC environment will be kept at the sufficient capacity and performance level
- achieved by collaboration with ITS
- partly funded by ITS directly, partly from the IIAI financial contributions from colleges
- Access to a well-designed and well-maintained deep learning/GPU/AI environment
- Access to problem-solving expertise
- Space in CoE available to hold such research-problem specific consultations
- Staff members will be available to
- discuss the problem
- state the problem in well-defined terms suitable for applying AI approaches
- discuss needs for training/evaluation/testing data
- propose a solution strategy
- help with obtaining pilot results
- help with writing a research proposal = writing the AI piece of a proposal (assumes that the staff member will become part of the proposal team with identified relevant % effort)
- Optimally, these staff members will be established researchers in this area with % effort to spend on this endeavor
- may be postdocs partly paid from research projects, partly from IIAI
- may be faculty partly supported from IIAI
- may be faculty who invest in such activities simply to become part of the research proposal team
- will likely be a mix of the above
- Education for users
- Short courses will be developed and offered to existing and prospective users
- Some will be topic/environment-specific
- Others will be general and possibly offered in collaboration with UI3
- These courses need to be developed – developers will be paid from the available resources
- CoE GPU-enabled classroom will be available for user education activities when not used for CoE courses