“Where’s the Spoon?”
All the advanced algorithms and AI tech in the world won’t scale if your underlying data isn’t clean and well-governed
The most fun parts of my job as a management consultant specializing in Big Data involve walking around New York City and meeting incredibly interesting people. One such meeting recently found me sitting with a data management veteran and currently Strategic Data Architecture leader at a major global investment bank. During our conversation, he knocked me flat with a simple, entertaining reference that succinctly describes one of the key challenges in industry-based Big Data, machine learning and artificial intelligence efforts today.
Movie fans from the 1980’s may remember Eddie Murphy’s seminal 1988 work, Coming to America[https://www.imdb.com/title/tt0094898/ ]. The film is hilarious and holds-up surprisingly well over time. My contact’s specific reference was to one of the cut-scenes that run under the credits at the end of the movie. In this scene, Eddie Murphy, costumed as an elderly white gentleman, recounts a story about eating soup at a restaurant. Describing this in text would be excruciating, so please just watch the clip below, instead, and we’ll talk about what it means on the far side.
Where’s the spoon, indeed. I can’t think of a more poignant analogy for the current state of Big Data and Advanced Analytics. Here’s what it means to me:
Most of the interest in data management today revolves around the benefits it can deliver to the business. Appropriately so. Those benefits are the soup in this analogy. But the danger of a myopic focus on the benefits is that it can blind us to the underlying mechanics through which those are engendered and supported. Those mechanics are the spoon. The waiter in Murphy’s story, equating to our own business stakeholders, is frantically targeting attributes of the benefit with his frustrated line of questioning [“too hot?” “too cold?”] and does not recognize that these bring him no closer to the answer he seeks. The restaurant patron, that’s us, keeps repeating the same refrain, “just taste the soup!” hoping to lead the waiter to a more holistic understanding of the situation. Only at the end, with the punchline, does the waiter recognize the underlying concern: NO SPOON. And, due to the journey he was forced to walk to get there, his understanding is deep, visceral, and part of cinematic history.
For me, this reference beautifully illustrates the role of governance in our current data management environment. And I mean “governance” writ large, including all that goes in to aggregating, normalizing, creating domain-specific semantic meaning around, securing, and providing carefully-managed access to enterprise data. Business leaders, newly energized about the offensive potential of data in their revenue planning [see also: https://medium.com/@drewharteveld/weaponizing-governance-303940b40035 ], pound the technology and data organizations for answers about why their machine learning and artificial intelligence programs aren’t delivering the volume of insights that they had expected. Meanwhile, the professionals who run those functions, many of whom have been wrestling with the octopus of enterprise data for decades, try desperately to lead those stakeholders to an understanding that it’s the underlying governance that is the limiting factor precluding expected scale.
Spoon problems are more easily solved than soup problems, anyway
The fantastic news is that, like in Murphy’s sketch, spoon problems are more easily solved than soup problems, anyway! The complexities behind brewing the perfect pot of soup [ML, AI, and the fast-changing techniques of data science] are many and troubleshooting them a daunting prospect. Meanwhile, the process of fetching a spoon [adopting and supporting sound and holistic data governance practices] is clearly understood and executed.
The moral is this: if you aren’t gaining the value you expect from your machine learning and artificial intelligence initiatives, consider stepping back to examine the enterprise data that underlies those programs. The data is the common denominator, and without being cocooned in holistic governance will continually foil your efforts to scale new techniques and technologies into the future.
“Whatta you know from funny, ya bastard.”
