The AI Playbook: Mastering the Rare Art of Machine Learning Development
Eric Siegel
MIT Press (February 2024)
“Vision without execution is hallucination.” Thomas Edison
Here in a single volume is probably what leaders of most organizations need to know about HOW to deploy machine learning throughout all levels and within all areas of the given enterprise. The value of the information, insights, and counsel that Eric Siegel provides is incalculable. As I worked by way through The AI Playbook, I was again reminded of an observation that Eppie Lederer (aka Ann Landers) made years ago: “If you think education is expensive, try ignorance.” Of course deployment poses unique and formidable challenges. Without effective collaboration with AI, success is probably impossible or at least highly problematic.
Siegel draws upon wide and deep personal experience when sharing his observations and recommendations. These are among the passages of greatest interest and value to me, also listed to indicate the nature and scope of Siegel’s coverage:
o Foreword (Pages xi-xiv)
o Preface (xv-xix)
o AI (3-4, 3-4, 82-83, and 82-83)
o When Greatness Is Too Big to See Up Close (5-6)
o A Tale of Two Technologies (18-20)
o BizML and BizML Cheat Sheet (21-22, 31-32, 46-47, and 95)
o The Two Main Technical Steps of Machine Learning (23-27)
o Many Models Never Deploy: An Industry-Wide Problem (27-29, 45-46, and 204-205)
o The Solution: BizML (31-32)
o The Semi-Technical Background Knowledge You Need (42-44)
o Proactively Preventing Bad Outcomes (70-73, 110-111, and 160-161)
o Establish the evaluation metrics (81-91)
o To deploy ML Is to Triage and Prioritize (94-96)
o Big Training Data for Ad Targeting (120-122, 166-168, 185-186)
o Train the model (141-147)
o A Summary of Modeling Methods (154-157)
o Shift Happens (170-172)
o When You Don’t Need Humans in the Loop (175-178)
o Mitigating Employment Risk with a Control Group (190-192)
o Conclusion: Morality Matters (209-212)
These are Eric Siegel’s concluding thoughts: “As you follow this book’s practice to get ML successfully deployed, make sure you are putting this powerful technology to good use. If you optimize only for a single objective such as improved profit, there will be fall-out and dire ramifications. But if you adopt humanistic objectives as well, science can help you achieve them.”
The AI Playbook is a brilliant achievement. In fact, I view it as a “must read” for all senior-level executives in almost all organizations, whatever their size and nature may be. Also, for those who are preparing for a career in business or have only recently embarked on one. Bravo!
* * *
Here are two suggestions while you are reading The AI Playbook: First, highlight key passages. Also, perhaps in a notebook kept near-at-hand (e.g. Apica Premium C.D. Notebook A5), record your comments, questions, and action steps (preferably with deadlines). Pay special attention to the boxed mini-commentaries that are strategically inserted throughout the lively and eloquent narrative (Pages 21, 63, 81, 106-107, 113, 141, 157, 165-167, 169, 195, and 197).
These two simple tactics — highlighting and documenting — will expedite frequent reviews of key material later.