“It’s Tough to Make Predictions, Especially About the Future”* (of the Internet)

Elon University’s Imaging the Internet Center aims to provide insights into emerging network innovations, global development, dynamics, diffusion, and governance. For over a decade, they have been collaborating with the Pew Research center to conduct regular expert surveys to support predictions in these areas.  

Experts responding to the last survey, conducted in 2014, yielded over twenty themes for the next decade, including:

  • “The spread of the Internet will enhance global connectivity that fosters more planetary relationships and less ignorance.”
  • “The Internet of Things, artificial intelligence, and big data will make people more aware of their world and their own behavior.”
  • “The spread of the ‘Ubernet’ will diminish the meaning of borders, and new ‘nations’ of those with shared interests may emerge and exist beyond the capacity of current nation-states to control. “
  • “Dangerous divides between haves and have-nots may expand, resulting in resentment and possible violence.”
  • “Abuses and abusers will ‘evolve and scale.”
  • “Most people are not yet noticing the profound changes today’s communications networks are already bringing about; these networks will be even more disruptive in the future.”


The next wave of the survey is underway, and I was asked to contribute predictions for change in the next decade, as an expert respondent.  I look forward to seeing the cumulative results of the survey, which should emerge next year. In the interim, I share my formative responses to questions about the next decade:

… the next decade of public discourse… will public discourse online become more or less shaped by bad actors?

The design of current social media systems is heavily influenced by a funding model based on advertisement revenue. Consequences of this have been that these systems emphasize “viral” communication that allows a single communicator to reach a large but interested audience, and devalue privacy, but are not designed to enable large scale collaboration and discourse.  

While the advertising model remains firmly in place, there hasbeen increasing public attention to privacy and to the potential for manipulating attitudes enabled by algorithmic curation.  Yet, I am optimistic. I am optimistic that in the next decade social media systems will give participants more authentic control over sharing their information and will begin to facilitate deliberation at scale.

… the next decade of online education, and credentialing… which skills will be most difficult to teach at scale?

Over the last fifteen years we have seen increasing success in making open course content available, followed by success teaching classes on-line at scale (e.g. Coursera, EdX). The next part of this progression will be online credentialing — last year, Starbucks’s partnership with ASU to provide long numbers of its employees with the opportunity to earn a full degree online is indicative of this shift.

Progress in on-line credentialing will be slower than progress in online delivery, because of the need to comply with or modify regulation, establish reputation, and overcome entrenched institutional interests in residential education. Notwithstanding, I am optimistic we will see substantial progress in the next decade — including more rigorous and widely accepted competency-based credentialing.

Given the increased rate of technical change, and the regular disruptions this creates in established industries — the most important skills for workforces in developed countries are those that support adaptability, and which enable workers to engage with new technologies (and especially information and communication technologies) and to effectively collaborate in different organizational structures.

While specific technical skills are well-fitted to a self-directed experience, some important skills — particularly metacognition, collaboration, and “soft” (emotional/social intelligences) skills  — which are particularly important for long-term success, require individualized guidance, (currently) a human instructor in the loop, and the opportunity to interact richly with other learners.

… the next decade of algorithms —  will the net overall effect be positive or negative?

Algorithms are defined essentially as mathematical tools designed to solve problems. Generally, improvements in problem solving tools — especially in the mathematical and computational fields have yielded huge benefits in science, technology, and health, and will most likely continue to do so.

The key policy question is really how we will choose to hold government and corporate actors responsible for the choices that they delegate to algorithms.  There is increasing understanding that each choice of algorithms embody a specific set of choices over what criteria are important to “solving” a problem, and what can be ignored. To incent better choices in algorithms will likely require actors using them to provide more transparency, to explicitly design algorithms with privacy and fairness in mind, and to holding actors who use algorithms meaningfully responsible for their consequences.

… the next decade of trust —  will people disengage with social networks, the internet, and the Internet of Things?

It appears very likely that because of network effects people’s general use of these systems will continue to increase — whether or not the systems themselves actually become more trustworthy. The value of online markets (etc.) are often a growing function of their size — which creates to a form of natural monopoly making these systems increasingly valuable, ubiquitous, and unavoidable.

The trustworthiness of this systems remains in doubt. It could be greatly improved by providing users with more transparency, control, and accountability over such systems. Technologies such as secure multi-party computing, functional encryption, unalterable blockchain ledgers, and differential privacy have great potential to strengthen systems — but so far the incentives to deploy them at wide scale are missing.

Structurally, many of the same forces will drive this as drive use of online networks. It appears very likely that because of network effects people’s general use of connected devices will continue to increase — whether or not the systems themselves are actually become more trustworthy.

The network of IoT is at an earlier stage than that of social networks — and there is less immediate value returned, and not yet a dominant “network” of these devices. It may take some time for a valuable network to emerge, and so the incentives to use IoT seem so far small for the end-consumer, while the security issues loom large, given the current lack of attention to systematic security engineering in design and implementation of these systems. (The lack of visibility of security reduces the incentives for such design). However, it seems likely that within the next decade the value of connected devices will become sufficient to drive people to use, regardless of the security risks, which may remain serious, but are often less immediately visible.

Reflecting on my own answers, I suspect I am reacting more as a “hedgehog” than as a “fox” — and thus am quite likely to be wrong (on this, see Phillip Tetlock’s excellent book on Expert Political Judgements).  I will in my defense recall the phrase that, physicist Dennis Gabor once wrote, “we cannot predict the future, but we can invent it.” — this is very much in the spirit of MIT. And as we argue in Information Wants Someone Else to Pay for It, the future of the scholarly communications and the information commons will be a happier one if libraries take their part in inventing it.

* This quote is most often attributed to Yogi Berra, but he denied it, at least in e-mail correspondence with me in 1997. It has also been attributed (with disputation) to Woody Allen,, Niels Bohr, Vint Cerf, Winston Churchill, Confucius, Disreali [sic], Freeman Dyson, Cecil B. Demille, Albert Einstein, Enrico Fermi, Edgar R. Fiedler, Bob Fourer, Sam Goldwyn, Allan Lamport, Groucho Marx, Dan Quayle, George Bernard Shaw, Casey Stengel, Will Rogers, M. Taub, Mark Twain, Kerr L. White.

See also: drmaltman