In the late 1960s, when I was an intern for the Department of Health, Education and Welfare, the department was working on a project analyzing the value of a college education. But there was a problem: The researchers didn't have a computer with the 64 kilobytes of memory needed to run the calculations.

Back then, 64K of computer memory was considered massive. Today, the typical smart watch has a memory 8,000 times larger than that. Processor productivity also has increased exponentially, doubling around every 18 months. Processors today are some 8.6 billion times more productive than those of the 1960s, and 75 percent of that power has been created in just the last three years.

These rapid advances in the capabilities of digital tools pose important challenges for the public sector -- challenges for governance that are often dangerously ignored. These are challenges that will loom larger as we put all that growing processor productivity to work harvesting new power through not only machine learning and artificial intelligence but also innovations such as quantum computing and blockchain infrastructures.

Digital innovation today needs broadly based and better governance not only to derive more productivity from government's value chains but also to protect and improve equity (fairness in the distribution of value) and trust (belief that the governance authority applied is legitimate). To build these new systems of governance, it's important to understand how digital tools work with these chains. They add value in three major ways:

Feedback and analysis, which clarify and measure relationships between inputs and outputs: Our 1960s HEW work suggested that community colleges were a good investment for increasing the lifetime earnings of students. Today, that kind of analysis is far less costly, much more powerful and available to use in more settings by more people. In one of many recent examples, web-based communications enable inexpensive testing of alternative texts to increase revenues from messages sent out to collect taxes and other bills.

Automation, which delegates tasks once performed by humans to computers: After many decades of work, for example, we have quite recently succeeded in having computers accurately translate text to voice and voice to text. And in just the last few years, many tasks once considered "uniquely human" have been computerized, such as identifying people from their pictures; assembling relevant documents on legal, medical and other subjects; and driving cars. The potential for computerization to fundamentally disrupt work and jobs is new and explosive.

Scale, which can decrease unit costs as production volume increases: Cloud technologies have greatly improved the cost-effectiveness of services standardized for large-scale, even global delivery. Digital "network effects" (where benefits rise with the number of participants served) are coming more into play, as governments such as Oakland County, Mich., share their data centers and applications with other jurisdictions. While technology is again the enabler, success depends even more on standardized approaches born from collaboration among independent institutions and jurisdictions. We're reallocating work within increasingly large value chains and networks.

Only a few decades ago, when technology was expensive, applications focused on small changes within relatively small groups of people. The hard part was getting the technology to work. For government, the response was to delegate projects to technology specialists and vendors. Today, however, the hard part is changing human behavior and jobs. In this setting, there is a much more serious threat in delegating IT to the techies.

In some settings we've begun to address these strategic challenges more aggressively, as, for example, in our evolving approach to cybersecurity. We have worked on application areas such as education, health care and public safety, in some cases not just as technology problems but as strategic concerns for entire industries and jurisdictions and their related professional and regulatory communities.

But what we are currently doing is simply not enough. With digital power poised for fundamental and disruptive impacts on work and life, we need to address the bigger and broader issues of today's full value chain, not just the smaller pieces we focused on in the past. Technology leaders will remain essential, but we need even more to educate and engage government's managers, workers, suppliers and policymakers, along with the citizens they serve.