I have been thinking a lot lately about the ubiquity of digital technology. The evolution over the past 70 years has been dramatic: batch computing in the 1950s, time-sharing in the 1960s, networking in the 1970s, the Internet in the 1980s, the World Wide Web in the 1990s, smartphones in the 2000s, social media in the 2010s, and now artificial intelligence (AI) in the 2020s. Concurrently, digital technology has become embedded in virtually everything, and that has implications for reliability, resilience, and repair. If anything goes wrong with a digital chip, it is often cheaper or even only possible to replace, not repair. Physical and electrical damage can be permanent. Try listing things that don’t have chips in them! Most appliances have chips for display and control. When these don’t work, there is often no recourse.
I wonder how many chips are in the Tesla I drive? When the screen display isn’t working, I can still drive, but I have only limited control over the car’s myriad functions. How about the elevators that ask you to select a floor, and then the system directs you to specific elevator? Once you are in the elevator, you are at the mercy of someone else’s software; there are no second choices. Washing machines, dishwashers, microwave ovens, COVID testing devices, wrist watches, television sets, heating and air conditioning systems, glucose monitoring and insulin injection systems, hearing aids, cochlear implants, pacemakers, vacuum cleaners, toilets, refrigerators, security systems, clocks . . . even my spaghetti fork has a timer in it! And I haven’t even gotten to 4G/5G/6G and Wi-Fi along with laptops, pads, and smartphones.
Increasingly, many of these devices are Internet-enabled, increasing their functionality and utility but also leaving us helpless if the network isn’t available. Artificial intelligence is adding another layer of functionality but also potential risk if it is not working as intended. These systems are stunning in their capabilities and equally surprising when they fail in unexpected ways. It is no wonder we are seeing a significant uptick in interest in what I will generally call “digital governance.” The United Nations entities have established two AI panels; in fact, at the time of this writing (in late August 2025), I was scheduled to meet with the Leadership Panel and the Multistakeholder Advisory Group of the Internet Governance Forum at the UN. Multiple Undersecretaries, the UN Secretary General, and scores of UN agencies have interests in digital technology, its capabilities, risks, and implications for society. Governments around the world are wrestling with human rights, safety, trust, and security in the context of an increasingly digital environment.
Computer scientists, engineers, and programmers are key players in this digital and online environment. Their work, or should I say our work, is central to the utility of these technologies and their benefits and potential deficits. They expect these complex artifacts to work as intended, and I think we owe it to them to deliver on that expectation. By implication, that means we need to work hard to ensure the resilience and reliability of these systems. Because the digital world is rife with dependencies, we are also responsible for understanding them and seeking to reduce the potential brittleness of the system. Some of us have started to ponder whether the Internet could ever be cold-started after a total shut down (like a Carrington-level EMP event—look it up). What information must be where to cold-start, and how does it get there? It really isn’t entirely clear. Similar questions are asked and sometimes successfully answered. When did you last hard-reboot your laptop or mobile smartphone? We owe it to an increasingly digitally dependent society to make its digital infrastructure as resilient, reliable, and safe as possible.
Join the Discussion (0)
Become a Member or Sign In to Post a Comment