Electronic spreadsheets have been around for nearly 40 years now. They were invented by Bob Frankston and Dan Bricklin, founders of VisiCalc, and I had a chance to chat with both gentlemen a couple of months ago. I highly recommend watching this TED talk with Dan Bricklin:
It’s important to understand for which purpose electronic spreadsheets were built in the first place if we want to anticipate what their future might look like. In response to my question in how far the future of spreadsheets is different from its past, Bob Frankston replied: “Not very different — smarter paper.”
Do spreadsheets have a future at all?
I remember when back in the 1980s the advent of the PC raised hopes for the paperless office. What happened in the next decades to come? As you might guess: the exact opposite. The corporate world’s appetite for paper grew beyond anyone’s expectations. There is a realistic chance that the same thing might happen with electronic spreadsheets, even though competition is on the horizon.
Programming languages such as Python are becoming increasingly popular in the corporate world, spreading far beyond the IT department. JPMorgan just recently made headlines with their attempt to teach their financial analysts to code. The bank presently employs 250,000 people, and the number of future Pythonists goes likely into thousands. Will initiatives like these erase electronic spreadsheets from the surface of corporate laptops and desktops? Probably not.
“To a large extent, Excel (and the like) democratized programming by making its power available to non-experts to solve their problems on their own. What will be the Excel of machine learning? Chances are it might just be … Excel. It makes sense to collocate the AI with the data.” François Chollet, Artificial Intelligence Researcher
François Chollet is the creator of Keras, the leading deep learning framework for Python. Due to his LinkedIn profile, Keras has around 250,000 users and 700 open-source contributors worldwide. Research labs such as CERN, Microsoft Research and NASA are utilizing Keras as well as many large tech companies like Netflix, Uber, and Google.
It does not take much imagination to picture what might happen if technologies such as Keras spread beyond the aforementioned sophisticated tech giants (aka early adopters) into the brick-and-mortar part of the corporate world: An ever-growing amount of data will be metabolized by AI-powered systems and ultimately spit out into … Excel.
Is this potentially the end of the world? Will humanity be taken down by a spreadsheet typing error?
“Microsoft’s Excel Might Be The Most Dangerous Software On The Planet” Forbes – February 13, 2013
In 2012, shortly before the above article was published, JPMorgan lost nearly 7 billion USD due to a trading error caused, to a certain extent, by miscalculations in Excel spreadsheets. Fast forward a decade, is Excel becoming even more dangerous due to the advent of AI?
Is Excel maybe even a ludicrously dangerous software?
Select from those three speed options when flying with Lord Dark Helmet:
- Lightspeed (too slow)
- Ridiculous speed
- Ludicrous speed
Very often, the fear of a catastrophe has a bigger impact on our lives than the catastrophe itself, which might never happen anyways. Humor has the power to release tension and clear the air, thus enabling an emotionally unbiased search for realistic solutions:
How can we make better use of spreadsheets in an AI-powered world?
In the aforementioned 2013 Forbes-piece, its author Tim Worstall quotes HITC:
“Both the Switzerland-based Basel Committee on Banking Supervision1 (BCBS) and the Financial Services Authority2 (FSA) in the UK have recently made it clear that when relying on manual processes, desktop applications or key internal data flow systems such as spreadsheets, banks and insurers should have effective controls in place that are consistently applied to manage risks around incorrect, false or even fraudulent data. The citation by the BCBS is the first time that spreadsheet management has ever been specifically referenced at such a high level, a watermark in the approach to spreadsheet risk.”
According to various resources, 80-90% of all spreadsheets contain errors. How to avoid them? ICAEW, the Institute of Chartered Accountants in England and Wales, with over 150,000 members, published a breathtakingly well-thought piece on this subject: “Twenty principles for good spreadsheet practice”
As Michael Izza explains in the preface:
“The Principles are the culmination of several months’ work carried out under the auspices of the Excel Community Advisory Committee, which brings together a group of 17 expert Excel users with vast experience in a variety of roles across business and practices, large and small.”
Their brilliantly clearly outlined principles are grouped into three categories:
- The spreadsheets business environment
- Designing and building your spreadsheet
- Spreadsheet risks and controls
ICAEW also does compliance checks:
“The IT Faculty has developed a scheme whereby spreadsheet standards, and other products and services such as training, can be formally recognized as compliant with the Twenty Principles.”
Finance and accounting experts are the silent early tech adopters
I am a corporate sales & marketing manager by trade who transitioned towards data science. That being said, my view on this subject is colored by my 25 years of professional experience: “How can a spreadsheet policy be sold into the lines of business instead of just being forced upon them?”
By all means, I do not suggest that spreadsheet policies should be optional. For obvious reasons, they must be mandatory. However, they have to appeal to early adopters specifically. Based on my experience with corporate spreadsheet users, those working in finance are the ones with the highest interest in enhancing the way they work with data, even up to the highest ranks. I met three CFOs who, independently from each other:
- learned to code
- acquired a fundamental understanding of machine learning and deep learning
- successfully applied what they learned in their finance department and beyond
Once a finance department cheerfully embraces spreadsheet policies, other departments likely will follow. This will not only make everyone’s lives easier and decrease their exposure to severe career risks, but it will open doors to successfully apply AI to the benefit ofh their employers.
I run a data literacy consultancy, and we cater to large companies in Europe and the US. You have questions I didn’t answer in my write-up, or you want to share your experiences? Please leave a comment or reach out to me via email [email protected] or LinkedIn. Thank you!