The end of digitalization?

Image by 00luvicecream from Pixabay

The last weeks, I traveled quite a bit (even ran a marathon!) and visited several of companies as well as talked to a lot of people in both industry and academia. Having had a few days to reflect on my experiences, I can’t shake an overall feeling that many are giving up on some of the key principles of digitalization. So, I decided to write an intermezzo post, in the middle of my Agile to Radical series, to do a bit of deep dive on this.

The pattern that I suspect I am seeing is the following: about a decade ago, leaders in many companies realized that digitalization, i.e. the use of software, data and AI, was going to affect them and their industry. Digital technologies allow for different forms of value delivery, e.g. continuous, fast feedback cycles using DevOps and data as well as automation of tasks that were unsolvable until the introduction of machine learning and specifically deep learning.

Many companies started to prepare for that future. Especially in embedded systems companies, this meant bringing software development in-house, hiring data analysts and AI experts and kicking off a bunch of experiments and prototype development to prepare for inclusion in future products and offerings. 

The results of these experiments and prototypes are now sufficiently mature for productization and this is where the wheels come off. Interestingly, it is all three digital technologies that run into issues when we try to bring these to products.

Especially in products that have a safety certification requirement, many companies realize that they have no idea how to conduct safety certification in a DevOps context. Traditionally, companies would do a one-off, big bang safety certification that would, worst case, take months to complete and then avoid touching the system unless it was unavoidable. Now we aim to bring DevOps to our offerings, but we fail due to the fact that we can’t or don’t know how certify in a recurring and cost-effective manner. And now the EU is in the process of mandating a new law around product liability which makes everyone in industry even more risk averse as we don’t know how this law will affect companies once it is in effect.

Then there is everything around data. All the legislation around GDPR, the lack of clarity in many companies around ownership of the data and end-user license agreements that fail to regulate everything related to data is causing a situation where companies basically don’t dare to collect, process and store any data that originates from systems that reside at customers.

To exacerbate the situation, the EU Data Act will force companies to open up the data that they do collect to third parties who can then build new and differentiating solutions on top of the rapidly commoditizing mechanics and electronics heavy offerings that many of the companies that I work with provide to their customers. This is causing companies to proactively reduce the amount and types of data they do collect in order to protect themselves from the competitive pressure that new entrants might provide.

Of course, the third digital technology, artificial intelligence, has had an enormous amount of attention. Many of the fear-mongers paint a picture of Skynet taking over and exterminating humanity. I do not want to repeat all the arguments in Marc Andreesen’s excellent piece on why AI will save the world, but I do want to mention that most of the people that are highly critical of AI and very vocal about it are “bootleggers”: they have ulterior motives and are continuously pushing the fear buttons that the general public is so sensitive to.

And, once again, the EU is coming to the rescue of the bootleggers with its AI act, causing a level of compliance overhead that will cause most companies to shy away from using machine and deep-learning in their offerings due to the legal and compliance complexities.

So, where does this leave us? If the input that I have received in the last weeks is at all accurate, I believe that especially European companies are withdrawing from adopting digital technologies and returning to the old way of doing things. OEMs in automotive are significantly slowing down their DevOps and autonomous drive roadmaps. Companies in industry and manufacturing have stopped collecting and processing any data from their customers and are instead pushing everything to the customer’s premises. Companies in healthcare are, by and large, waiting for the FDA as well its international counterparts, to tell them how they can work with DevOps, use patient data and deploy machine learning models.

The standard argument in this context is that we are doing this to avoid unintended, negative consequences. But nobody talks about the cost of slowing technology adoption down. According to the world health organization, 1.35 million people die every year due to traffic related accidents. Autonomous driving solutions deployed worldwide would reduce this to a fraction of that number. Who owns the responsibility for all the deaths that could have been avoided if we had moved faster on adopting autonomous driving solutions?

Similarly, instrumenting equipment in factories and continuously improving performance has enormous potential for reduced waste, increased efficiency and reduced risk for the people working in these factories. Who is responsible for waste, reduced efficiency and accidents due to factories not incorporating these digital technologies?

And do we even have to raise healthcare in this context? If there is one industry where hardworking and incredibly well-intended people are overburdened by administrative and regulatory overhead it is healthcare. According to one study I found, about 25% of all cost in healthcare are due to this overhead. The expanded use of digital technologies on the administrative side and even more on the diagnosis and treatment side of healthcare would save or at least extend millions of lives. Who do we blame for all the people that die prematurely or live lives with severely reduced quality due to the slow adoption of new technology?

Many describe me as a hard-core modernist who simply believes that technology can do no harm. Of course, I am not that naive. Technologies ranging from dynamite to face recognition software can be used for bad purposes. However, the vast majority of technology uses is positive for humankind. From the adoption of agriculture, the industrial revolution, the computer revolution, big data and now the AI revolution, every time we bring a new technology to bear, the benefits for humanity are immense. Of course, it brings change and disruption, but we end up in a better place. Virtually all of us work so hard on these things to make sure that our children have a better life than what we have and technology, and currently especially digital technology, is the most promising avenue for progress.

For all the attempts at regulation, history teaches us that humans suck at predicting the impact of new technologies. So, we are regulating the wrong things and are unnecessarily slowing down technology progress. On the other hand, humans excel at responding to challenges that materialize. Once a threat is real, we take care of things swiftly and effectively. So, rather than trying to dream up a host of nightmare scenarios, let’s work on adopting the technologies as rapidly as possible and then fixing the places where things go off the rails.

Concluding, I see a reversal of the pendulum and many companies, as well as society at large, moving away from or at least slowing down the adoption of digital technologies. The regulatory burden and public opinion concerning new technology is so challenging that many companies, especially in Europe, are slowing down innovation and we are stuck in a status quo that is maintained as long as possible in a, seemingly evil, pact between regulators, e.g. the EU, and incumbents who use this to reduce competitive pressures. I am, at heart, a positive and constructive person, but I am really worried and concerned about these developments. My ask to you is to do what you can, in your circles of control and influence, to change things for the better. As Nelson Mandela said, it is in your hands to make a difference.

Want to read more like this? Sign up for my newsletter at or follow me on, LinkedIn ( or Twitter (@JanBosch).