The technology crystal ball

“Only a fool, a charlatan or, possibly, the Chancellor of the Exchequer claims to be able to predict what will happen in the future..."

The only thing anyone seems to predict with certainty anymore is how uncertain we all are about the future. This is very frustrating because we always seem to end up paying for it, whether in accrued tech debt or surprise downtime, because something somewhere in the chain that makes up modern technology has collapsed.

At the time of writing (see, even the author isn’t sure if it’ll still be around in a few months!), ChatGPT has burst onto the stage. This has gone from an R&D chatbot in progress to show off the new features of the GPT model, too, according to some, a vital part of their tools to get work done. And slowly, perhaps too slowly, people are asking the question: wait, what about security? Data protection and privacy? Who owns the data anyway? Is this even sustainable long term? We want new technology to remain competitive companies need it, but without answers to security, privacy, responsibility, sustainability and stability concerns, we can never fully trust these technologies. This has had impacts, too, with the rise of challenger banks like Monzo offering a mobile app. Even the notability risk-averse banking system has had to compete and catch up. The question becomes, can we take a peak into our technology crystal ball, predict what technologies are up and coming and then focus on developing security, stability, privacy, responsibility and trust early on in the hype cycle before it places people at risk? And this is a big deal. How do you balance the rise of technological progress against the security of not just a single individual but an entire society?

“Those who cannot remember the past are condemned to repeat it.”.

What about security? Well, we don’t have an answer. But we’re working on that. If you are interested in participating, we’d love to interview you for the next part of this study! We’re looking for individuals interested in technology not in the next year or five years but in technology in 2040.?Please sign up here, and if you are selected, we will follow up within 30 days to invite you to an interview.

We should start not at a look towards the future but instead towards the past. The study of the future is called Future Studies or, perhaps more commonly recently, Futurology. And it’s not all sci-fi and plucking ideas from thin air! Many approaches are used to predict the future: From examining social media to conflicting discussions as experts come to a consensus. Each method has surprising results; perhaps we are not as bad at telling the future as we first appear to be [1]. So how on earth do we predict the future with any kind of accuracy?

Social Network Analysis

Social network analysis is about understanding the communities and circumstances around an idea. A good way of thinking about this is on social media websites, where users can post an opinion, and people can reply. Social media platforms encourage engagement, so those users who post their opinions, engage in discussions in their community and share that discussion are particularly valuable to the platform, ensuring users continue to spend their time on it. However, these relationships can be positive, negative, or neutral and have different meanings. Trends can be revealed by mapping out this complex web of interactions and relationships. This could be how positive or negative a particular reply is to an idea, how often individuals communicate, whether or not they are collaborating, and more subtle clues like the individual’s larger networks and how deeply connected they are to one another. Let’s look at a real example of this in action, programmers often collaborate with others to build new features or fix bugs, this takes a lot of coordination on both sides, and if they are not success can lead to the failure of a program to build (e.g. the code has a bug so it can’t even get turned into a program that people can run), communication is key [2]. To get work done, developers often use planning tools. These will track code changes, allow developers to report issues, comment on changes, tasks or issues and/or let individuals watch a discussion or subscribe to see new comments or tasks. These can be modelled as a mathematical graph, which each node being a contributor and each interaction is an edge or line between them. The graph can be analysed when a build fails to understand where the communication broke down. I’m sure this won’t be a surprise, not least to anyone reading who is in software. Still, it turns out a lack of communication in multiple types of available communication between contributors could predict the failure of the software to build.

Delphi: getting consensus

The Delphi method involves several rounds of interviews with experts until a consensus has been made. This is very labour-intensive, requiring multiple experts to spend their time speaking with researchers, but it provides some measure of what experts think. The challenge comes in finding those who are indeed experts and willing to spend their time talking to researchers. Let’s look at football to get an idea of what this looks like. Sports, in general, have seen both an increase in technology and seem somewhat untouched. There are no sci-fi mech suits to go beyond the limits of the human body; instead, very little has changed, with athletes using banned performance-enhancing drugs, and while the chemicals may be different, the effect remains the same. However, we have seen a remarkable increase in data usage, where coaches can understand their movement and play style and turn this into a prediction of the game’s victors. But we’re not interested in predicting the victors of a single match. We want to predict the technological trends of the next 10-20 years [3].

So step one of any prediction method is to figure out what we will predict by identifying projections from related work for football [4]. This leads us to 3 themes, player-related, coach-related and technical director related. We can then develop these into projections “In 2026, players are more involved and interested in data-driven decisions around their physical and tactical performance” this process uses a mix of consultations with expert interviews, starting with a general “how will technology impact…”. Participants can then select how desirable and impactful that particular projection will be, elaborating in an open text. In the second round, they can revise their projections informed by the results from the first. The results then become clusters based on probability, impact and desirability, noting any interesting results. For football, this is that it is unlikely for technology to impact the game. However, instead, it supports game-related decision-making or training decisions.

Scenario planning: going extreme

Scenario planning takes another approach; it starts by finding a driving force or a trend in eternal uncertainties not consistent with the assumptions in the scenarios. Scenario planning imagines possible future through a thorough objective evaluation of alternatives, what if then, for multiple criteria to inform discussions among stakeholders for a shared understanding of that uncertainty. For instance, in the energy sector, energy transition is a huge uncertainty [5]. Carbon is currently being replaced by solar and alternative energy sources, is there enough infrastructure and technology to match the scale of change/transition? How fast will it go? Is the pace of renewable development an equal match? Are the right technologies available to achieve the transition? What about security and energy poverty? And many more. These questions and more are explored in scenario planning to provide a meaningful framework to get a grip on these uncertainties to inform strategic decision-making for the future.

Oh, looking at the predictions from previous years concerning the future we are living in now, some predictions came true, and others did not. That’s why scenario planning for multiple, equally plausible versions of the future is a necessity.

Futures Wheel: understanding consequences

In contrast to some of the other methods we’ve looked at, which often start with a prediction or some idea of the answer, the futures wheel method takes a different approach. We start with an event, trend or idea in the centre, and then participants are asked to think about its consequences (intentional or otherwise). Then you repeat with the first set of consequences, creating a list of consequences of the consequences, and go again, with a list of consequences of consequences of consequences. The consequences aren’t all negative, including some positive and neutral consequences in addition to the negative. Once participants generate this wheel, they then discuss it, refinding ideas and coming to a conclusion.

Again, let’s look at an example, and because this is a blog post written after 2020, we’ll start with the defining event of the 2020s, the COVID-19 pandemic [6]. One of the consequences of a brand new strain of coronavirus was the closure of borders and travel bans. Because of this, airlines and the tourism industry were hit particularly badly, causing job losses, event cancellations, and supply chain issues. This allows governments to focus on recovering the tourism, travel and business-travel parts of the economy, with policies such as “Eat Out to Help Out” designed to offer discounts for hardest hit industries. This can also be used over a longer perspective when considering supply chain resilience. As the global microchip industry struggled during the pandemic, prices rose dramatically, and many businesses struggled to find the needed microchips for their products. Now we, as a society, look to evaluate our supply chains and build resilience as Europe faces a gas supply issue.

Are we good at predicting, then?

Well, let’s take a look! The seminal prediction was published in 1967 and was titled From The Year 2000: A Framework for Speculation on the Next Thirty-Three Years [7, p. 200]. It makes several predictions about society in the 2000s. While there are some hilariously wrong predictions, such as number 27, which predicts the use of nuclear weapons for excavation and mining and 99, which predicts the use of an artificial moon to light up areas at night, some are chillingly correct, particularly those involved in communications. Perhaps this is not surprising for a piece of work written within the context of the Soviet-American arms race and well before the collapse of the Soviet Union.

“27. Use of nuclear explosives for excavation and mining, generation of power, creation of high-temperature/high-pressure environments or for a source of neutrons or other radiation

99. Artificial moons and other methods of lighting large areas at night”

While it is important to avoid allowing these predictions to “cold-read” the 2023 reader, while written within the context of the 1960s, many communication and computing predictions were correct. For example, 75 predicts a time-sharing agreement for computers, which we know as cloud computing today. Similarly, 81 predicts personal “pagers", which a 2023 reader would recognise as modern smartphones.

81. Personal "pagers" (perhaps even two-way pocket phones) and other personal electronic equipment for communication, computing, and data-processing)

75. Shared-time (public and interconnected) computers generally avaliable to home and business on a metered basis

Interestingly this list was reviewed in 2002 [1], and even then seems out of date, as the first recognisable smartphone, the iPhone, would not release until 2007. Some of these are now possible even when looking at the worst forecasts (judged by 2002). For example, the effective ballistic missile defense, which a modern reader may know as Isreal’s Iron Dome, which has been in service since 2011. Or, human hibernation, which may be recognised as a medically induced coma. While in its earlier stages, individual flying platforms have also been developed using similar technology to drones [8].

From the original 1960s list, 81% of innovations within the theme of communications and computers were judged to have occurred by 2000, based on the 2002 review. This theme includes data processing, computers, networks, video, and additive manufacturing (3d printing). For other themes, the predictions are worse with defence, materials, biotech/agriculture and environment, with 50-40% accuracy. The worst theme was aerospace which primarily predicted interplanetary travel and habitation. While aerospace has developed significantly between 1967 (pre-moon landing) and 2002, even in 2022, permanent habitation outside earth still looks far away.

This leads us to the next question: If these are the themes for technological change in 1967, which for some were accurate, what are the themes in 2022? Unlike in 1967, in 2022 we have some major advantages, thanks to our “personal pagers”, “personal computers [...] to communicate with the outside world”, and “inexpensive high-capacity worldwide [...] communication [...] with light pipes”. We can judge and measure “hype” when a technology has an abundance of “hype”, many individuals, both in the scientific community and the general public, are excited to see what this technology will bring. However, this excitement does not always translate to the implementation; it can become “over-hyped” with a lot of interest and excitement without practical uses or benefits. These are often measured on a theoretical hype cycle; initially, a new technology has a lot of interest in a very short time, but as this interest grows, there are problems this could be practicality, implementation or commercialisation issues, leading to a steep decline, before these problems are mitigated, and the technology becomes a mainstay. Technologies that currently have a lot of hype the “personal pagers”, could become the smartphones of 2040.

A technology that has evolved from science fiction is the so-called Metaverse, virtual and augmented reality, while initially, we looked for holograms on spaceships, the power of smartphones has molded this into a dog Snapchat filter. Alongside this development has meant the development of AI, intelligent and autonomous systems, though not limited to the detection of faces, with models such as ChatGPT being available publicly. It is difficult to imagine a future without technology becoming more integrated into our lives. The connect places theme looks to smart cities and grids as part of this future, able to balance our energy needs by carefully computing the energy generation against energy needs. Finally, quantum and next-generation computing offer a future where computing is not as we know it today, using qbits to solve some of our biggest optimisation challenges and computing the future.

Bibliography

[1]???????R. E. Albright, ‘What can past technology forecasts tell us about the future?’,?Technol. Forecast. Soc. Change, vol. 69, no. 5, pp. 443–464, 2002, doi: https://doi.org/10.1016/S0040-1625(02)00186-5.

[2]???????T. Wolf, A. Schroter, D. Damian, and T. Nguyen, ‘Predicting build failures using social network analysis on developer communication’, in?2009 IEEE 31st international conference on software engineering, 2009, pp. 1–11. doi: 10.1109/ICSE.2009.5070503.

[3]???????D. Adam, ‘Science and the World Cup: how big data is transforming football’,?Nature, vol. 611, no. 7936, pp. 444–446, Nov. 2022, doi: 10.1038/d41586-022-03698-1.

[4]???????D. Beiderbeck, N. Evans, N. Frevel, and S. L. Schmidt, ‘The impact of technology on the future of football – A global Delphi study’,?Technol. Forecast. Soc. Change, vol. 187, p. 122186, 2023, doi: https://doi.org/10.1016/j.techfore.2022.122186.

[5]???????T. Witt, M. Dumeier, and J. Geldermann, ‘Combining scenario planning, energy system analysis, and multi-criteria analysis to develop and evaluate energy scenarios’,?J. Clean. Prod., vol. 242, p. 118414, Jan. 2020, doi: 10.1016/j.jclepro.2019.118414.

[6]???????P. Daffara, ‘Applying the futures wheel and macrohistory to the covid19 global pandemic’,?J. Futur. Stud., vol. 25, no. 2, pp. 35–48, 2020.

[7]???????A. J. Wiener and H. Kahn, ‘The Year 2000: A Framework for Speculation on the Next Thirty-Three Years’, Oct. 1967.

[8]???????P. Holderith, ‘The SkyPak Is a Prototype Electric Jetpack That Actually Works’,?The Drive, Apr. 07, 2022. https://www.thedrive.com/tech/45110/the-skypak-is-a-112-hp-drone-backpack-prototype-that-actually-works (accessed Feb. 22, 2023).