Discover more from All About Data
"Data Culture" hits the (digital) shelves!
A managerial yet very practical guide to putting Data and AI at the heart of large and small enterprises
Article originally in English; automatic translation into Italian available here.
As Nassim Taleb says:
History doesn’t crawl; it leaps.
I definitely believe we are witnessing some leaps that will extensively impact our future, when we think of the recent developments in the world of Data Science, Machine learning and Artificial Intelligence.
We’re talking about disruptive innovations that will have a great impact on the way we work and the way companies succeed… or fail. That’s the point: we tend to focus on the latest success, but the vast majority of Data and AI projects are failures that are quickly forgotten by the public. Or sometimes hidden from the public. How many of you remember Galactica, the large language model launched by Meta and suspended in just 3 days, one week before the launch of ChatGPT by OpenAI?
In statistics, this is called survivorship bias.
How does all of this relate to “Data Culture”, the book I recently published with Stefano Gatti? Let’s find out!
Two sides of the same coin
Stefano and I have been working together for the last 6-7 years, after very different journeys. He spent more than 10 years leading the Innovation and Data Sources area of one of the largest Italian (and European) credit information providers, spreading the Open Innovation paradigm and leading investments in promising startups. While I found my way in Data Science only after a few years in a different field (IT Security), becoming so far the only Competitions Grandmaster in my country on Kaggle, the leading platform for machine learning competitions, which was acquired by Google a few years ago.
What we share is a common passion for data and algorithms, especially when it comes to using them on real world problems in large companies.
And we independently came to the same conclusion: somehow, despite the amazing developments of the last few years, most enterprises are struggling heavily to really use data at its full potential, all over the world. The number of press releases, statements and conferences where companies claim to be data-driven is growing at the same pace of the difficulties that the same companies are actually facing when working with data.
Why? I think that we can rule out at least two options.
The technology (underlying AI) is racing at an unimaginable pace on every possible aspect. It may be difficult to keep up (that’s for sure), but there are many solutions for every possible need and innovative paradigms, like the Cloud, that are lowering the need for huge investments up front. And in order to get started, there are plenty of amazing technical books (or other resources), written by data experts for other data experts, that support the understanding of the latest technological advancement.
The vision (on how AI will shape our world) is supported by many texts and essays that provide great insights into what the future of artificial intelligence will look like. The slant of these books may be different, more geared toward the professional or the merely curious: in any case, there are many facets and even more interesting points of view that allow one to form an opinion.
So what is missing for a real diffusion of data and algorithms, not only in the Big Tech world, but also for more traditional small and large enterprises? Why is it reported in a famous Gartner article that 85 percent of AI projects fail in various ways and why are data scientists at the top of the list of professionals seeking new job opportunities, often leaving seemingly perfect jobs?
We believe that something is missing, and it is not the technology or the vision: it is the culture, that intangible but fundamental element that facilitates the execution of an effective data strategy.
What to expect from “Data Culture”
This is certainly not a philosophical book on the concept of data culture, and I don’t think any person could expect something like that from two engineers lent to writing, such as Stefano and myself.
If anything, the practical sense is the common denominator of the five chapters that make up our book.
Let's cut to the chase. In quick succession, here are the topics that we cover:
Software is eating the world - with data and algorithms. Let's frame the technologies of this world and see why they are not the starting point for those who want to make the best use of data.
Organization matters. Moving from the Chief Data Officer onto the other roles, we figure out who does what, who is needed more and who is needed less...
We often speak of data science as a single matter, but it is not. “Data scientists” working in research centers and “data scientists” employed by large traditional companies have less in common than you might think. Let’s take a look at what they actually do.
Those who know little about data think they can measure anything. Including the ability to use data to their best advantage. Those who are experts in measurements, on the other hand, are asking tough questions on how to evaluate the effectiveness of a data and AI team. Anyone familiar with the Dunning-Kruger effect will easily understand.
The future of data and algorithms is unknown. But with a bit of experience and the help of some friends, we try and analyze 10 trends, using our personal crystal ball to envision how they’ll be like in the next 10 years.
Who’s this book meant for
The natural readers of this book are all those professionals who in various capacities have to deal with data and algorithms, especially those who want to set out on a path of transformation and innovation, avoiding falling into the myriad traps and misunderstandings that are at the root of the large number of failures mentioned above. And into which, in some cases, we fell first, in our own journey.
But I actually believe that many others may find this reading interesting. Think of the famous Conway diagram:
“Data Culture” speaks not only to those in the intersection, but also to those who mainly belong to only one of the three fields:
Business experts and people/project managers
Mathematicians, statisticians and "quantitative" people of all types
In my experience, anyone who belongs to just one of these three categories sees data professionals as distant relatives: they look somehow familiar and not entirely unknown, but in actual terms little is known about their world. This book is a good opportunity to dig in and understand a little bit more!
Lastly, even those who have recently started working with data and AI (or those who aspire to enter this world) will find many useful and non-trivial insights in the book: every single page is influenced by many years of work in the field, and I think that a critical reading can be a window into the way of thinking of two people who have been working in this field for years.
It’s not (only) about the experiences of the authors… it’s about our guests!
I just wrote “two people”, but actually the book is enriched by contributions from data executives who we’ve been delighted to interview along the way.
Massimo Chiriatti, currently CTO in Lenovo Italy after a long career in IBM, has given us a very personal and incisive preface that reminds us that every company is unique because it has unique people and unique data.
The five chapters are concluded by as many interviews with data experts and professionals:
Carlo Torniai, literally a citizen of the world, who lead data scientists in multiple countries and companies, such as Tesla, Pirelli, FIFA, Esselunga, always with a keen eye on the intersection of data and business
Anna Russo, who applies data science and AI in the fashion industry (as Global Director of Data Science in Gucci), after years in Skyscanner, Channel 4, QuantumBlack
Sébastien Bratières, a Frenchman living in Rome and an engineer with a deep passion for languages, who is director of AI at Translated (a European scaleup working on NLP) and managing director at Pi School
Giovanni Paganini, an applied mathematician who started working with data and AI before it was cool! With experiences as an executive in IBM, SAS and EY.
Maria Parysz, a Polish entrepreneur who founded 3 startups, worked in 3 different continents, and somehow managed to create a physical twin to what was only an online community, Kaggle. By the way, a community of over 10 million data scientists, in case you didn’t know!
Last but not least, Abraham Thomas gave us an amazing afterword, at the same time far-reaching and down-to-earth. As the co-founder of Quandl (a pioneering company of alternative data that he sold to Nasdaq) and an angel investor with a priviledged perspective on the data and AI world, we thought he was the perfect match to give us a glimpse on the future.
So where can I find the book?
Well, if you’ve read this far and you're curious to go deeper, just head to Amazon! “Data Culture” is available in both ebook and paperback versions.
Here is the link: https://www.amazon.it/dp/B0BZTCN932
There are many other things to say, but I’m stopping here for now.