The AI Agile Manifesto

As I was preparing a talk about my new book on AI Project Management, The AI Project Handbook, I thought I would put the lessons from the book into one single document. I realized I was really trying to define some AI Agile principles. I call the document the AI Agile Manifesto in honor of the 2001 Manifesto for Agile Software Development (written twenty years ago!). Have a look and tell me what you think. I will revise it based on your comments and suggestions. 

The AI Agile Manifesto

  1. You need to know what you want to achieve with AI. There is a trade-off between feasibility and business impact.
  2. The organization has to be committed to the AI project.
  3. The AI team leader needs to be an effective manager and a leader with a shared vision of AI.
  4. Design thinking and Agile are valuable tools. Be laser-focused on the to-do-list to control the scope, cost, and schedule of the AI project.
  5. You need to know all the factors that can influence the AI project.
  6. The AI project needs to leverage and be consistent with all the organizational process assets.
  7. An AI project needs great people, models, and data.
  8. AI quality is not only about model and software quality but also about people and data.
  9. AI risk management requires a constant risk assessment, a risk strategy, and Human-in-the-loop.
  10. You need to involve all the stakeholders and have a clear communication plan, especially when things go wrong.
  11. You need to identify, understand and address the ethical concerns caused by AI.
  12. Agile for AI requires a specific approach with longer cycles and more exploration.                                                           

Leave a Reply

Your email address will not be published. Required fields are marked *