Effective Integration and Responsible Use of AI Platforms Into Day-to-Day Analytics EffortsGenentech
summary
Claims that average ROI from AI deployments takes about two to two and a half years and that trust varies by country (US ~40%, India highest, China second). Recommends combining LLMs/ChatGPT-style models with rule-based systems and domain review into hybrid analytics so sentiment analysis and taxonomy/classification work better. Emphasizes human in the loop, transparency about models (LLMs mostly predict next word and can regurgitate training data), strong governance and data quality; two different AI tools used on the same task produce noticeably different reformatting and reliability.
AI-written summary · about
chapters (10)
- 0:00AI and analytics integration
- 1:35AI and tech cycles
- 3:25Effective AI integration principles
- 6:43LLM limitations in analytics
- 9:13Predictions to insights
- 11:01Global AI adoption and regulation
- 13:09AI for people
- 14:17Hybrid analytics approaches
- 19:36AI implementation lessons
- 22:12Ethical AI and future
related talks
37mPanel: AI Priorities in the Enterprise
13mAI as a Strategic Asset: Building Competitive Advantage
48mPractical Strategies for Successful AI Adoption in your Organisation
01h 01mPanel discussion: Practical and responsible applications of AI in businesses
07mOne-on-One with Yusuf Jazkallah and Sam Gilstrap
30mHow organizations can put their data to work and build 10,000 responsible, human-centric AI models
26mAI in Use: Industry Perspectives
45mPanel: Rethinking Skills and Work in the Age of AI