Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
SlideShare a Scribd company logo

1

Australia’s National Science Agency
Liming Zhu
Research Director, CSIRO’s Data61
Conjoint Professor, UNSW
Generative AI for
Enterprises
All pencil drawings in this presentation are created by AI

2

Generative AI
•Text
•Image/video
•Code/Scripts
•Data
Predictive
Diagnostic
Concepts Clarified
2 |
Prescriptive
Discriminative
Model
Generative
Model
Hybrid

3

Generative AI Transformation - Business
3 |
• General Capability
• human resources or tools/functionality
• Ease of Access
• More low-cost experimentation driven
• Less problem/cost-benefits driven
• Blurring/Moving Division Boundaries
• reverse Conway's Law

4

Generative AI Transformation - Architecture
4 |
Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. Towards Responsible AI in the Era of ChatGPT: A Reference
Architecture for Designing Foundation Model-based AI Systems. https://arxiv.org/abs/2304.11090

5

Generative AI Transformation - Infrastructure
5 |
“Talk to Your Data” --> FM-based Innovation/Operational excellence
“Zero-gradient Infrastructure” --> FM tuning and FMOps
Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. A Framework for Designing
Foundation Model based Systems https://arxiv.org/abs/2305.05352v1

6

Responsible Generative AI
6 |
Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. Towards Responsible AI in the Era of ChatGPT: A Reference
Architecture for Designing Foundation Model-based AI Systems. http://arxiv.org/abs/2304.11090
Product/Services
Development Processes
Governance
https://research.csiro.au/ss/science/projects/responsible-ai-pattern-catalogue/
CSIRO Responsible AI
Pattern Catalogue

7

Potential Use Cases
7 |
Operation
• Automated reporting
• Digital Twin?
• Multi-model/instrument data
integration & synthesis
• Operation plan generation
• Enhanced public communication
Digital Subject Matter Experts (SME)
Innovation
• FM-based integration
• Generative AI surrogates for M&S
• Synthesis of experiment &
simulation data and code
• Rare/extreme event prediction
• Cross-discipline observation

More Related Content

Generative-AI-in-enterprise-20230615.pdf

  • 1. Australia’s National Science Agency Liming Zhu Research Director, CSIRO’s Data61 Conjoint Professor, UNSW Generative AI for Enterprises All pencil drawings in this presentation are created by AI
  • 3. Generative AI Transformation - Business 3 | • General Capability • human resources or tools/functionality • Ease of Access • More low-cost experimentation driven • Less problem/cost-benefits driven • Blurring/Moving Division Boundaries • reverse Conway's Law
  • 4. Generative AI Transformation - Architecture 4 | Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. Towards Responsible AI in the Era of ChatGPT: A Reference Architecture for Designing Foundation Model-based AI Systems. https://arxiv.org/abs/2304.11090
  • 5. Generative AI Transformation - Infrastructure 5 | “Talk to Your Data” --> FM-based Innovation/Operational excellence “Zero-gradient Infrastructure” --> FM tuning and FMOps Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. A Framework for Designing Foundation Model based Systems https://arxiv.org/abs/2305.05352v1
  • 6. Responsible Generative AI 6 | Lu, Q., Zhu, L., Xu, X., Xing, Z., Whittle, J., 2023. Towards Responsible AI in the Era of ChatGPT: A Reference Architecture for Designing Foundation Model-based AI Systems. http://arxiv.org/abs/2304.11090 Product/Services Development Processes Governance https://research.csiro.au/ss/science/projects/responsible-ai-pattern-catalogue/ CSIRO Responsible AI Pattern Catalogue
  • 7. Potential Use Cases 7 | Operation • Automated reporting • Digital Twin? • Multi-model/instrument data integration & synthesis • Operation plan generation • Enhanced public communication Digital Subject Matter Experts (SME) Innovation • FM-based integration • Generative AI surrogates for M&S • Synthesis of experiment & simulation data and code • Rare/extreme event prediction • Cross-discipline observation