Red Sky in the Morning

Red sky in the morning, sailors take warning.

Growing up I heard this old nautical adage that a red sky in the morning means foul weather is coming, while a red sky at night indicates fair weather on the horizon.¹ I wish predicting the future were so simple today.

Like many of us, I’ve spent time over the past few months reflecting on what to expect out of the next few years. My exploration has ranged from estimating trajectories for businesses and society, to positioning my work, family, and life accordingly. I also completed a course on the future of war, so my reflections had a slight national security bent. I’ve reproduced an excerpt below for the sake of reflection and discussion. I would love to hear your thoughts.

This analysis considers the future international operating environment. It begins with Current Strategy and Policy, an assessment that considers the latest U.S. national security strategy and national defense strategy to evaluate the current U.S. perspective on and plans for the future. Next, the Current Conceptualizations of the Future section evaluates the future trends for a period 15 to 30 years from now, based on research from the U.S. Joint Chiefs of Staff, U.K. Ministry of Defense, Atlantic Council, and the U.S. National Intelligence Council.

Major themes in the analysis of the future include:

• a shift in global power balance to a multipolar, fragmented, increasingly contested world;
• increasing significant (violent?) actions by both state and non-state actors;
• rapid changes in technology that increase asymmetric capabilities and open up new common operating arenas; and
• a general increase in overall uncertainty and instability.

The description of the Future Operating Environment Scenario picks up where the foresight research leaves off and highlights the effects of recent developments in artificial intelligence (AI) and automation, extrapolating forward to the 15- to 30-year period and adding expected robotic automations. In a follow-on work, I present a Recommended National Security Strategy that offers solutions to the issues identified in this analysis of the future operating environment.

Introduction

In order to consider changes to a strategy, it is important to understand the current intentions and the current assessment, as well as any predicted changes that will come in the future. With these three sets of data, the strategist can make a plan for the future.

This assessment begins an analysis for future national strategy by analyzing the current intentions based on the 2025 United States National Security Strategy. It then considers the expected future operating environment based on three governmental resources and one think tank report. From there, the analysis looks forward with other data and trends to inform the most likely changes to those future predictions. In particular, this analysis highlights recent developments in artificial intelligence and automation, considering a scenario where AI, automation, robots, and drones are rapidly deployed and useful at scale, leading to wealth inequality and societal disruption, governance challenges, and an international system struggling to cope. The analysis works through assessments of the future environments in technology, governance, the international system, and conflict.

Current Strategy and Policy

The 2025 United States National Security Strategy is the most complete contemporary view of U.S. national strategy. It is organized into the goals the U.S. should have, both for its internal actions and what it desires from relations with other countries, as well as the means it has to achieve those goals. It is additionally broken down into the principles and priorities of that strategy and the regions of the world. For this review, relevant aspects of the strategy will be organized by topics that fit under technology, governance, the international system, and conflict, in order maintain a common structure.

Technology

The current U.S. National Security Strategy follows the usual trend of emphasis on developing and fielding the most advanced technology as a primary competitive advantage. In particular, the document highlights a desire that “U.S. technology and U.S. standards—particularly in AI, biotech, and quantum computing—drive the world forward.”² It also calls out a need to invest in research and development to ensure “continued prosperity, competitive advantage, and military dominance for future generations.”² In particular, it highlights efforts in areas where the U.S. already has an advantage, including “undersea, space, and nuclear” as well as future tools like “AI, quantum computing, autonomous systems.”² It likewise highlights the importance of pursuing the energy to fuel future technologies² and the importance of a resilient national infrastructure.²

Governance

The National Security Strategy calls out the goal of “full control” over the borders and immigration system² as well as the “restoration and reinvigoration of American spiritual and cultural health…”² It highlights a need for the “world’s strongest, most dynamic, most innovative, and most advanced economy”², the “world’s most robust industrial base”², and the “world’s most robust, productive, and innovative energy sector.”² The strategy notes that a strong economy is the “bedrock of our global position and the necessary foundation of our military.”²

The International System

The document shares the goal of having the “most powerful, lethal, and technologically advanced military” while maintaining ‘unrivalled “soft power.”’² It highlights a renewed focus on the U.S. place in and leadership of the Western Hemisphere², as well as a shift to a “Burden Sharing and Burden-Shifting”² and “commercial diplomacy” model of international relations.² It outlays a “Win the Economic Future, Prevent Military Confrontation” strategy in Asia.² Notably, the strategy highlights that the “world’s fundamental political unit is and will remain the nation-state.”²

Conflict

As mentioned, the National Security Strategy carries the goal of having a powerful and technologically advanced military. It also highlights non-military threats, such as “espionage, predatory trade practices, drug and human trafficking, destructive propaganda and influence operations, [and] cultural subversion.”² It also calls out economic security as “fundamental to national security” and highlights the following areas of emphasis: balanced trade, securing access to critical supply chains and materials, reindustrialization, reviving the defense industrial base, energy dominance, and financial sector dominance.²

Current Conceptualizations of the Future

There are several resources from both governments and think tanks that provide research and analysis of future trends and developments. Each is organized differently, but a summary of important concepts is offered here, again separated into technology, governance, the international system, and conflict.

Technology

In most of the foresight documents, technology is seen as a strategic equalizer and a potential destabilizer. Most of the documents view technological change as potentially accelerating faster than institutions can adapt.³ In particular, artificial intelligence, autonomy, cyber capabilities, space systems, and biotechnology are reshaping power.⁴ This reshaping is expected to favor speed, data, and adaptability over scale. It is also expected that this development in technology, in particular at a fast pace, will raise inequality from the deployment of robotics, artificial intelligence, and automation.⁵ Specifically, the Atlantic Council predicts of the future: “Technology increases inequalities within and between states.”^{5 6}

Governance

The foresight documents identify several stressors on global governance. First, the research shows expected demographic divergence, with the potential for social strain throughout the world.^{4 5} Many advanced economies will see aging and shrinking populations, which could lead to financial stress and social welfare burden on the younger working populations.^{4 5} Additionally, in fast-growing regions such as Africa and South Asia, rapid population growth and urbanization is predicted, which could lead to raised youth unemployment and political volatility.⁶ Each of these trends is expected to lead to rising migration pressures that test social cohesion and governance.

Additionally, populations that are more educated, connected, and demanding could see socio-political expectations grow faster than governments can deliver.^{6 3} Such a gap could fuel protests, polarization, and declining institutional trust. Spurred by increased connectivity via the internet, there is a potential for a reduction in national identity and citizenship as individuals emphasize self and group identities in their lives.⁵

The International System

Most of the foresight documents agree that the international system in the future will include a more contested and fragmented international order.^{4 5} In particular, the post-Cold War order is expected to give way to a multipolar contested system. International power will be more diffuse, with rising states—but also cities, corporations, networks, and non-state actors—becoming increasingly important on the world stage.^{4 7} Additionally, the rules, norms, and institutions of previously important actors will be increasingly ignored, selectively applied, or actively contested.⁴ The foresight research expects an increase in grey zone operations below the threshold of open war, not only by traditional smaller non-state actors but also by larger state actors. These operations could include increased coercion amongst actors using economic pressures, cyber operations, lawfare, and information warfare.⁴

Conflict

As mentioned, the international order of the future is expected to be more contested and more fragmented. Smaller and weaker actors are expected to gain outsized influence using new technologies such as in the cyber domain, drones, and artificial intelligence. For many, stresses from the climate and environment will serve as a core security driver.⁶ In general, there will likely be persistent disorder or compound crises from pandemics, financial shocks, climate disasters, technological disruptions, and geopolitical crises that increasingly overlap.^{4 6} Governments may find themselves in states of near-constant crisis management. For these states, strategic depth would depend on resilience, redundancy, and adaptability. Indeed, national power will be measured less by static indicators and more by the ability to absorb shocks, learn quickly, and adapt faster than rivals.⁶

Future Operating Environment Scenario

In a review and assessment of the United States National Intelligence Council’s (NIC) previous Global Trends reports prior to beginning work on Global Trends 2030, independent analysts concluded that the NIC was successful in “identifying the key drivers of change” and offered quality regional analysis. However, they noted that the previous analyses frequently fell short in forecasting the rate of change, underestimating the actual pace.⁵ The following is an assessment of the world and future conflict that is focused on recent major developments in artificial intelligence (AI) and the effects that are expected to come from rapid advancement over the next 15 to 30 years. The assessment considers the following major themes, which will be described in greater detail in the following sections.

First, AI and robotics serve as a means to convert electricity to productivity. It follows that improved electricity generation and management leads to more power, more wealth, and increased ability for individuals, companies, and countries that do well as they leverage artificial intelligence and robotics. This enhanced importance for energy could also lead to a race for nuclear fusion or any other generation of electricity, increased use of solar panels, wind turbines, gas turbines, as well as potentially even locating compute resources in space.

The battlefield of the future is nearly all drone warfare. Future drone swarms are a complete overmatch against humans. The true weapon in this future operating context is the production system for these autonomous fighting drones. Given the drone-human overmatch, these machines would fight not on national sovereign territory, but instead in the global commons, from the oceans, polar regions, and skies above to cyberspace, orbits, and potentially the lunar surface.^{8 4} The edge formed between combatant drone swarms will serve as a shifting front line in conflict, fluctuating as production on each side varies higher or lower and as the availability of energy resources fluctuates. This warfare of the future is a constant balance of swarming techniques, controls, and drone concentration on a target with changing disposability metrics based on national drone production and replacement rates.

The governance and social impacts of this newfound technology, automation, and efficiency will show as increased economic inequality, reduced social cohesion, and increased fragmentation in most societies.⁸ International order becomes multipolar as countries band together for AI expertise and energy imperialism. Smaller states not yet involved with a larger power may struggle to choose an alignment or struggle to act as one entity, given an increase in populations finding identity outside of the traditional Westphalian nation-states.

Other issues that are certainly important and relevant to this future timeframe, but not in the immediate focus given these other critical pressing concerns, are climate change, aging populations or related changing demographics, and urbanization. The next several sections will look to the above effects as they appear in technology, governance within states, the international system between states, and global conflict.

Technology

When considering the expected rate of change in technological advances over the next 15 to 30 years, it is helpful to look back at change over the previous similar period. Since 1995, changes in technology show impressive development and growth, with consequential impacts on individual lives, companies, and countries. In the technology arena, the wireless standard for Wi-Fi began in 1999.⁹ Residential broadband cable was first available in 2002, but it was not until 2010 that two-thirds of homes reported having it.¹⁰ The iPhone was only released in 2007, less than 20 years ago.¹¹ In the same year, Netflix launched streaming services.¹² Both of these technologies were supported by public cloud computing resources, the beginning of which can be estimated with the Amazon S3 offering in 2006.¹³ Facebook launched in only 2004.¹⁴

Machine learning has been around since the 1950s, but the relatively recent availability of large amounts of data and powerful compute resources led to the 2012 NeurIPS conference paper on AlexNet, a deep learning breakthrough on using convolutional neural nets (CNNs).¹⁵ Google published the landmark paper on transformers, Attention is All You Need, in 2017, and OpenAI released ChatGPT in 2022.^{16 17}

In energy, the combination of hydraulic fracturing with horizontal drilling began the U.S. shale revolution in the early 2000s, and the cost of utility-scale solar photovoltaics fell by about 85 percent from 2010 to 2020, making it obviously cost-competitive at scale.¹⁸ These examples for the rate of change in digital, machine learning, and energy technologies in the recent past can help frame the future scope and rate of technological change.

It is expected that AI capability will gradually increase, one day surpassing humans at nearly every task. The change may be steady but feel sudden, such that people could be caught by surprise as one day their jobs and livelihoods are secure, while the next day they are not. In addition to frontier model development, the improvement process could come from innovations involving teams of models or agents, each tailored to a specific task, and leveraging hundreds or thousands of them together to replace a standard human worker. Compute hardware could run three or four of these human-equivalent agentic teams at full power continuously. After the initial purchase and setup, the only significant cost would be the energy to power them. More electricity applied toward a process would mean more agent teams, higher digital productivity, and better results.

On the hardware side, the competitive advantage shifts from the details of the product being sold to the automation of the manufacturing process that makes it. The fastest-growing companies will have automated factories that build automated production line robots that eventually produce the products sold to humans. The best-performing companies will hit a positive feedback loop of production speed once automated factories build more factories or software builds more software.

In this world where computers turn electricity into productivity, energy drives everything. The new currency will be electrons. Companies with more energy can run more agentic teams. Individuals with more electricity can run more models, use more agents, and achieve better results. Everything will become a function of energy. In science, having more energy will lead to leveraging more compute, which will lead to more novel discoveries. For medical diagnoses and analyses, using more energy will allow for higher scan power or a higher number of scans, more compute put toward the analysis, and a higher quality diagnosis. There is a world where an MRI or a CAT scan could have three levels of quality—low, medium, and high—all related to the amount of energy being used and therefore the total cost.

Raw materials will still be a limiting factor in building physical products, such as goods, physical infrastructure, and robots. Some synthetic production could be possible, but it will likely be easier to have mining robots gather the materials. It would not be surprising to see a return of resource-based imperialism in the future by technologically-advanced, automation-focused countries like the United States and China.

Even the nature of thinking will have changed for individuals—the human brain will no longer need to remember details or work through certain logical processes. Rather, the most productive brains will be master coordinators of agentic AI and organizers of the output data.

In this system, wealth inequality will grow between companies, countries, and individuals. Excess returns from this capital allocation, energy management, and AI automation will circle back to the entities that had excess resources or capital to deploy initially. Those with an advantage will see their advantage grow with a digitally- and AI-enabled positive feedback loop.

Governance

As AI technology is improved, the flywheel of capitalism will spin faster, challenging traditional governance. Individuals and companies with excess resources will invest in further leverage. They will use more AI, see greater returns, and earn more excess resources (profit). These competitors will be incentivized to use these resources to improve automation workflows and spin the flywheel even faster. This could lead to: the disappearance of most white-collar desk jobs, the automation of many manufacturing jobs, stable service work reserved for people with connections, some individuals in various populations fleeing to subsistence farms to escape this reality, and potentially many others in need of government support.¹⁹ The result of the above flywheel effect—especially combined with other contemporary trends and pressures—will be a rise in economic and wealth inequality, a reduction in social cohesion, a rise in fragmentation, a reduction or shifting in national identities, and increased polarization.^{8 4}

The traditionally-oriented nation-states could see a “declining legitimacy of state authority”⁴ and associated decline in state sovereignty, control, and influence.⁸ In a socioeconomic system like this, traditional democratic processes may be seen as moving too slowly and failing to build a system that works for all.⁶ The existing government structure could be too slow to adapt to major techno-socioeconomic changes and overly focused on short political election timeframes. The combined result of this could be a rise in autocracy and a rise in populism, as well as a general pushback away from liberal democracy and the globalized system.

There will likely be some strife among populations as previously agreed-upon social contracts will need to be renegotiated, such as with social security and other entitlement programs in the United States.⁸ Traditional schooling and training pipelines might become quickly outdated, leaving many people with obsolete skills up against state-of-the-art computer-powered competition. The traditional markers of success like education and corporate status might soon prove irrelevant, which could strain existing social structures. Similar issues could arise as individuals or populations struggle to define their purpose or roles in society.⁸ In The Coming Wave, Suleyman writes about this pending technological change: “The delicate bargain of the nation-state will be placed under immense strain just when we need institutions like it most.”¹⁹ Upstart revolutionary movements may kindle periodically and law enforcement drones would be tasked with quelling them.

There will be more below on the consequences of such a decline in a states’ authority within its own population. For now, it is worth noting that the communications, misinformation, and influence tools that could exacerbate polarization or a decline in legitimacy already exist. Competitor nations and peoples will leverage these and future tools to compete. States that lack authority with their own people will likely also lack authority and trust on the world stage, which could have negative ramifications for those countries.

The International System

In the future international system, the countries most advanced in artificial intelligence will take the lead and the overall system dynamics will see a tension between multipolarity and fragmentation. As AI development continues, the quality of productive output will follow model scale or size, then algorithmic efficiency—or creative thinking modes such as reasoning, compute efficiency, reinforcement learning, or continuous learning. Some notable portion of advancement will come from improvements in hardware efficiency. Lastly, AI performance will be a function of pure energy input, or amount of electric power.

In most of these areas, the United States and China have started off ahead of other countries. However, as they continue developing more digital and physical automation, they will need more energy and raw material resources—a form of energy imperialism will have begun and natural resource imperialism restarted. Countries like the United States and China, who have a lead in AI, will target countries with the above resources. Those smaller countries will have the option of a forceful takeover, or they will willingly allow mining robots to extract their natural resources and collect energy. Most of these smaller countries would be as concerned about other foreign major powers impeding in a similar way, so they will more-or-less willingly align with the power most aligned with their worldviews.⁶

As competition between countries grows and the world moves from unipolar to multipolar, China likely competes by closing down the South China Sea to shipping of goods or materials and continually increasing its influence on Southeast Asia, the Middle East, and Africa, as they had already started with the Belt and Road Initiative. The United States will move to maintain control of the Western Hemisphere, spreading its influence north and south into Canada, the Caribbean, Central America, and South America, all the while pushing back on the remnants of the People’s Republic of China (PRC) Belt and Road influence in South America. Russia, the European Union, and the United Kingdom compete for power over their areas to the north but generally avoid US and Chinese attention due to their lack of easily available resources.

In all, the international system will trend away from a unipolar world toward multipolarity, with the continued rise of China and a plateau and gradual decline of the United States in relative international importance. The role of non-state actors, such as multinational corporations, will increase as their influence spreads beyond that of the traditional Westphalian nation-state.^{20 8} Global companies will have increasing power, influence, and control, especially in those areas that are opened by technology but where national governments are unable to adapt quickly enough to gain power, such as the global commons: the oceans, polar regions, cyberspace, space, and the lunar surface.⁸ Massive companies with global investors and customers, may be incentivized to push back against the desires of a single country or group of countries. Throughout this evolution, there will be a gradual decline of globalization, and the rise of autarky (self-sufficiency), autocracy, and authoritarianism to combat the reduction in state control and sovereignty.^{8 6}

As the influence of national governments wanes, political and fiscal uncertainty will start to affect the global supply and demand for national treasury bonds or financial instruments. A negative effect could be exacerbated with higher government debt loads, either from overleveraged normal operations or from increased government support to populations struggling from the effects of automation. This will be especially true in cases where other countries or large global corporations offer higher returns with moderate or comparatively lower risk. Global funding would shift to maximize the perceived reward-risk ratio, and United States government bonds may be increasingly less desirable. From this trend, we would also likely see a gradual shift to multipolar reserve currencies, away from primarily the U.S. dollar.⁶

On the home front, having reduced control of bond rates, a government could struggle to maintain control of its monetary policy and fiscal liabilities. The lack of government control could exacerbate the loss of confidence from its population, adding to a positive feed loop of political and fiscal uncertainty for the country. Other countries with the means to do so could begin seeing opportunities to affect such a nations’ monetary policy or fiscal outcomes, especially by leveraging global markets that are hyper-connected and can change in real time. Financial operations could be wielded more directly as a tool of national power, hybrid warfare, or unrestricted warfare.

Conflict

International competition in the future, or future conflict, will be significantly less kinetic and utilize more of the various other tools of national power. It will align better with the concepts of hybrid warfare or unrestricted warfare. Countries will continue to maintain military power, but other tools of national power, such as informational, economic, financial, and intelligence tools, will see a growing importance.

When future conflict is kinetic, it will be nearly fully autonomous with swarming, low-cost drones. These drones will be in the air, on and below the sea, on land, in space, and include digital automation in cyberspace. Generally speaking, humans will be less capable, more costly, and less reliable than automated weapons. In this situation, the competitive advantage for a country will come from, first, the control and swarming technologies, and second, the country’s ability to replace individual units as they are lost, or the production lines. As this evolution progresses, humans will first want machines to arrive on station, load, make ready, and aim a weapon. Then, the machine would require authorization from the human before firing.

At some point in the technological development of these weapons, the human-in-the-loop process will not be fast enough, and humans will nearly never have to override the firing operations. At this point, the model of drone warfare will evolve to defining a goal, programming rules of engagement, and relying on the intelligent machines to optimize for success. Issues from this process would likely prove rare, and if they do occur, they would probably be considered collateral damage. As mentioned before, most conflict would not involve or be near humans anyway, as it would occur primarily between machines in the global commons.

In other competitive situations, people will redefine what “conflict” means. In particular, this will be a transition gradually toward a broader “competition,” where the competitive space is expanded beyond military force into using other tools of national power and aligning with the idea of either hybrid warfare or unrestricted warfare. Examples of this trend are above, including disruptive economic and financial measures. Additional efforts could take new forms like forced social disruption, polarization, or disinformation of a population using global information networks and social media influence.

On the international stage, a country’s power of attraction or its international “customer value proposition” will vastly increase in importance and must be maintained through means other than purely military power. Overall, there will be a notable change in the perception of international conflict from envisioning weapons, drones, missiles, and militaries today to considering how countries will actually win against each other in the future: leveraging better profitability, financial stability, political stability, resource gathering, resource ownership, supply chain stability, economic productivity, and generally providing value to their own people and those of other nations.

Conclusion

The future operating environment will be characterized by newfound speed and efficiency from artificial intelligence, machine learning, and robotic automation. This innovation will result in a flywheel of capitalism that spins faster than ever before, and governments must be able to react quickly to support their citizenry.

On the world stage, countries must prove that they are capable of this rapid response, both to ensure internal stability and to offer value abroad. Countries will begin moving away from a primarily military preparation toward competing in various domains, from economic and financial to military efforts in new battlespaces. As they do this, the previous unipolar world will gradually dissolve and nations will need to work together to compete, leaving a world with multipolar tendencies. Additionally, the race for energy to power various forms of digital or physical automation will color international competition and success. Whichever country comes into a great source of electricity generation—like nuclear fusion, for example—will likely lead the world in productivity for the foreseeable future. Given the uncertainties in this fast-moving future age, innovation and adaptation will prove exceedingly valuable.

— The next article lays out a plan for staying ahead: Red Sky at Night — Positioning for the Future

Footnotes

1. Library of Congress, “Is the Old Adage ‘Red Sky at Night, Sailors’ Delight. Red Sky in Morning, Sailors’ Warning’ True?” https://www.loc.gov/everyday-mysteries/meteorology-climatology/item/is-the-old-adage-red-sky-at-night-sailors-delight-red-sky-in-morning-sailors-warning-true-or-is-it-just-an-old-wives-tale/ ↩

2. US Office of the President, National Security Strategy of the United States of America (Washington, DC: Office of the President, 2025). https://www.whitehouse.gov/wp-content/uploads/2025/12/2025-National-Security-Strategy.pdf ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹

3. US National Intelligence Council, Global Trends 2040: A More Contested World (Washington, DC, 2021). https://www.dni.gov/files/images/globalTrends/GT2040/GlobalTrends_2040_for_web1.pdf ↩ ↩²

4. US Joint Chiefs of Staff, Joint Operating Environment 2035 (Arlington, VA, 2016). https://apps.dtic.mil/sti/pdfs/AD1012885.pdf ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹

5. Mathew Burrows, Global Risks 2035: The Search for a New Normal (Washington, DC: Atlantic Council, 2016). https://www.atlanticcouncil.org/in-depth-research-reports/report/global-risks-2035/ ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

6. UK Ministry of Defence, Global Strategic Trends 7: Out to 2055 (2024). https://www.gov.uk/government/publications/global-strategic-trends-out-to-2055 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰

7. Qiao Liang and Wang Xiangsui, Unrestricted Warfare (Naples, Italy: Albatross Publishers, 2020). ↩

8. UK Ministry of Defence, Global Strategic Trends 6: The Future Starts Today (2018). https://www.gov.uk/government/publications/global-strategic-trends ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

9. IEEE Standards Association, IEEE Standard for Information Technology, IEEE 802.11b (1999). ↩

10. Pew Research Center, “Home Broadband Adoption,” various survey years, 2008-2010, https://www.pewresearch.org/internet/fact-sheet/internet-broadband/. ↩

11. Apple Inc., “Apple Reinvents the Phone with iPhone,” press release, January 9, 2007, https://www.apple.com/newsroom/2007/01/09Apple-Reinvents-the-Phone-with-iPhone/. ↩

12. Quentin Hardy, “Netflix to Deliver Movies to the PC,” New York Times, January 16, 2007, https://www.nytimes.com/2007/01/16/technology/16netflix.html. ↩

13. Amazon Web Services, “Amazon S3 Announcement,” March 14, 2006, https://aws.amazon.com/about-aws/whats-new/2006/03/13/announcing-amazon-s3---simple-storage-service/. ↩

14. Facebook, “Company Info,” https://about.meta.com/company-info/. ↩

15. Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton, “ImageNet Classification with Deep Convolutional Neural Networks,” in Advances in Neural Information Processing Systems 25 (2012). https://proceedings.neurips.cc/paper_files/paper/2012/file/c399862d3b9d6b76c8436e924a68c45b-Paper.pdf ↩

16. Ashish Vaswani et al., “Attention Is All You Need,” in Advances in Neural Information Processing Systems 30 (2017). https://arxiv.org/abs/1706.03762 ↩

17. OpenAI, “Introducing ChatGPT,” November 30, 2022, https://openai.com/index/chatgpt/. ↩

18. International Renewable Energy Agency, Renewable Power Generation Costs in 2020 (Abu Dhabi: IRENA, 2021), https://www.irena.org/publications/2021/Jun/Renewable-Power-Costs-in-2020. ↩

19. Mustafa Suleyman, The Coming Wave: AI, Power, and Our Future (New York, NY: Crown, 2023). [] ↩ ↩²

20. Frank Hoffman, Conflict in the 21st Century: The Rise of Hybrid Wars (Arlington, VA: Potomac Institute for Policy Studies, 2007). https://www.potomacinstitute.org/images/stories/publications/potomac_hybridwar_0108.pdf ↩