Projectified How to Avoid AI Project Failure
12 snips
Mar 26, 2025 
Kathleen Walch and Ron Schmelzer, experts in AI project management at PMI Cognilytica, explore the pitfalls of AI projects. They discuss common failure reasons like poor data quality and the unique challenges of managing AI. The duo emphasizes the need for structured approaches and setting smart metrics to ensure real ROI. They also share strategies on balancing urgency with quality and the importance of continuous learning in the rapidly evolving AI landscape. Tune in for insights on navigating the complexities of successful AI initiatives!
AI Snips
Chapters
Transcript
Episode notes
LLM Hallucinations
- Large language models (LLMs) often "hallucinate," producing confident yet incorrect answers.
- This stems from scope issues; models perform best within their training data and context.
Problem-Focused AI
- Focus on solving real, valuable problems with AI, even if they're not massive.
- Identify which aspects of the problem require AI and determine the appropriate AI patterns.
AI vs. Traditional IT Projects
- AI projects differ from traditional IT projects as they are data-dependent and not solely functionality-driven.
- Changes in data significantly impact AI system performance, requiring a different project management approach.