Expert: Andy Uttley, Co-Head Business Data, Analytics & Tools at Capital Group Facilitator: Leigh Fisher, Head of Business Development, EMEA, Door
Headlines:
- Compared to broad AI, machine learning and deep learning, generative AI goes beyond pattern recognition to create new content
- A key framework is the 'expectations cycle' showing the path from innovation trigger to inflated expectations, disillusionment, enlightenment and productivity
- Capital Group takes an enterprise approach to AI with oversight by a Steering Committee driving speed and an AI Risk Oversight group ensuring control
- Its AI project backlog is maintained with prioritized use cases assigned across client group, investment group and enterprise functions
Context:
Key topics covered include differences between AI and generative AI, Capital Group's enterprise AI approach, AI oversight governance model, use cases across client group, investment group and enterprise functions, demonstrations of Copilot and video translation models, and philosophical questions on AI displacing jobs.
AI use cases including Copilot for knowledge retrieval, video translation, text translations, earnings call sentiment analysis, investment thesis tracking, analyst research summarization and news personalization were demonstrated, before attendees shared their own AI adoption experiences and ideas.
In an open discussion, attendees debate AI's impact on jobs, need for human judgment, behaviour changes from AI tracking, compliance risks, data privacy issues and strategic questions around AI investment and job displacement planning.
Key takeaways:
- Test and adopt AI models like Copilot for knowledge retrieval, video translation and text generation based on accuracy, speed and data privacy assessments
- Define tolerance thresholds for AI accuracy across different use cases such as client -facing communications, internal research processes, etc
- Review AI outputs for cultural nuances and limitations around interpreting gestures or sentiments
- Assess long-term workforce impacts and re-skilling needs from increasing AI automation and augmentation
