Move Fast and Break Things worked for high flying Silicon Valley startup megatech companies but it doesn’t work for AI projects. Between 70%-80% of AI projects are failing to meet their objectives. With stats like this, it's clear that breaking things isn't leading to success in AI. In this episode of AI Today hosts Kathleen Walch and Ron Schmelzer dig into this topic.

Iterating to Success vs. Breaking Things

AI projects often fail because organizations are tempted to abandon them when immediate success isn't achieved. Instead of moving fast, organizations need to focus on iterative sprints that bring AI projects closer to their goals. The key is to "think big, start small, and iterate often." This allows for incremental progress and small wins. In this episode we explain what iteration looks like for AI projects.

The Real-World AI Disconnect

Certainly, on this podcast we talk a lot about proof-of-concept versus pilots. AI projects frequently fail when small POCs are pushed into real-world applications too quickly. The disconnect between controlled environments and real-world complexities often leads to project failures. And we have seen this far too often.

Additionally, in this episode we discuss that AI projects must account for real-world data and conditions to succeed. We highlight the importance of aligning development environments with actual use cases. We explain why a cautious, step-wise approach to AI projects, even for simple tasks like document classification using NLP, ensures that AI solutions are tested and refined in real-world scenarios. A popular and proven approach is the Cognitive Project Management for AI (CPMAI). It emphasizes the importance of iterative progress and realistic expectations.

Show Notes:

• Free Intro to CPMAI course









• CPMAI Certification









• Subscribe to Cognilytica newsletter on LinkedIn









• The Necessary (and often Missing) “U” in the DIKUW Pyramid [AI Today Podcast]