Digital Innovation - When Backwards becomes the New Forward

09 / 06 / 2022

When we strive for digital innovation as humans, we have two options: using tools or getting others to help us. How does this change with current advances in computer science?

computer woman

When the recent Japanese deputy chief of cybersecurity was asked if nuclear facilities used USB sticks, Mr. Sakurada professed that he did not know what that was

“Since the age of 25, I have instructed my employees and secretaries, so I don’t use computers myself.”

The Mercury Seven were used to computers who’d listen very attentively to their safety concerns and respond with precision, a full explanation and perhaps a reassuring smile to the most complex computational task. In fact, Alan Shepard wouldn’t commit to any rocket trajectory that had not been verified by the one computer he trusted most, Mrs. Katherine Johnson.

Whenever we find ourselves at the limits of our capabilities, we naturally search for tools or helpers to empower us. And there the key challenge is the continuous effort of gauging what we need to invest into tool-skills or helper-relationships vs. the utility they provide, and whether an increasingly symbiotic relationship with either will retain long term benefits.

Mechanical machinery and computerized automation have picked up an ever-increasing share of hard, rote, fully engineered and formalized jobs, yet enabled human minds to create an ever greater and more complex range of things and services: payment is no longer exchanging a couple of chickens in return for a knife and whether cryptographic computations deserve more intrinsic trust than government issued paper in exchange for gold, has become only one of many aspects in a financial transaction.

Because the human mind may have far fewer limitations on its ingeniousness than it has on the number of complex individual people or services it can deal with—as carefully explained by Robin Dunbar[1]—the only road to obtain value from this ever-wider range of persona is to collect and abstract them into a minimal set, each of which can cover a major area of responsibility or life’s aspects.

But whether you are a commissioner of the EU, a corporate CEO, run a pizza delivery service, a nation’s military or a FinTech, the number of personae you interact with during most of your working day cannot scale with the number of people your actions directly affect or the budget involved, because the human brain power envelope is narrowly perched on a sugary metabolism that needs constant sustenance.

To increase your reach, value, empowerment or convenience requires the type of intelligence that bright, discrete and loyal assistants deliver by (a) observing your every input, output and action, (b) forming their theories of your mind which they (c) confirm or disprove with a minimum of discrete questions, (d) applying their ever-increasing knowledge base to performing ever more responsible tasks, which they (e) execute with greater degrees of autonomy and fewer checkpoints, (f) fully employing sub agents to scale their tasks to the size required to serve your purpose.

That type of skill completely exceeded what computers could do until recently, so the Japanese minister was probably right to let them be the tools of his assistants.

Advances in machine learning and artificial intelligence made possible via step changes in computational power have enabled a mixed paradigm, where bookkeeping chores still remain with your ERP system, but where artificial assistants become smart enough to learn routine tasks that hitherto only humans could master.

Their talents won’t be the same, common sense is still in much shorter supply on computers, even after having absorbed all of Wikipedia and textbooks on economics. But the never slacking attention to detail and the ability to potentially look at vastly more context in seconds than hordes of humans could peruse in hours can give them a significant functional and financial advantage.

Organizations which adopt these emerging hyper-automation products and services should soon outpace competitors who don’t; but of course, they’ll need to adapt and learn how to use this technology, which will definitely not match Mrs. Johnson’s reassuring smile, but hopefully reach her skills and justify a similar confidence.

For far more details on the technical and political challenges around that I recommend you dive into HiPEACs Vision 2020, which you can find here:


[1] Dunbar. 2010. How Many Friends Does One Person Need?: Dunbar's Number and Other Evolutionary Quirks. London: Faber & Faber ISBN 978-0571253432

Thomas Hoberg

Technical Director, Worldine Labs