Search the archive.

In 2018, we moved closer to the multipolar world that looks set to replace the bipolar US-Russian geopolitical regime that emerged from the Cold War. China’s ascent as a serious economic and geostrategic rival for the US, and its growing assertiveness with programs like "One Belt, One Road" or "Made in China 2025", has strengthened its influence on the world stage. The possibility that the US might withdraw from its nuclear arms treaty with Russia (China was not a signatory and therefore does not face the same limits) is a further testimony to the shifting geopolitics. This also holds true for historical alliances. Take the US and Europe, for example, which admittedly already constituted two close but distinct economic and strategic poles. Trade and defence issues have recently revealed additional cracks in this partnership, evidenced by their diverging positions on the Iran nuclear deal and sanctions. Struggling Western middle class Today’s multipolar world mirrors the vast development of emerging markets, fuelled by supportive demographics and the benefits that economic globalization brings. At the same time, it also reflects the relative decline of the middle class in developed markets. Since the 2008 financial crisis, inequalities have grown, not so much across as within countries, especially in developed countries. The economic policy mix of fiscal austerity and very loose monetary policy, which most developed economies adopted to respond to the financial crisis, proved particularly detrimental to the Western middle class. According to a 2015 report by the Pew Research Center, 61% of US adults lived in middle-income households in 1971, with the remaining 39% living in upper and lower-income households. In 2015, the two groups were equal at 50% each. Encroaching protectionism Tough labour market conditions, marked by persistently high unemployment and stagnating incomes, followed the financial crisis. These were exacerbated by hyper-globalization and disruptive technologies. This combination left many middle-class households with the sense of having lost out the most in the ensuing years. Their growing frustration led voters in the US, along with those in many European countries, to drive a political shift away from the establishment to populist governments with a new mandate to restore Western middle-class prosperity and security. As a result, economic policy in the West has become tilted toward more fiscal policy and protectionism. The newly elected governments’ key objective is to support their domestic economy through job creation and wage increases at home, and potential regulation or taxation of certain sectors or industries. In developed markets, for example, this takes the form of more restrictions for foreign companies seeking access to domestic markets and/or greater support for domestic companies to raise their competitiveness, for instance via corporate tax cuts. Beyond the US, where the Trump administration has taken action on both of these fronts, even countries like Germany, which were traditionally open to free trade, have started to prevent acquisitions by foreign companies when strategic knowhow is involved. The blocked acquisition of Leifeld Metal Spinning AG by Yantai Taihai Group is a case in point. As for emerging markets, there is increased awareness that they are the consumer markets of the future due to the sheer size of their population. Governmental growth policy in these markets is moving to a consumer and services-led model, with growing median private wealth predominantly benefiting domestic companies. Should the recent trade truce between China and the US shift away from tariffs and other barriers as key policy tools, corporate and private tax competition could become the new battleground. National champions take on new role Against this backdrop, national champions and brands take on an entirely new role. National champions are companies in large countries with a sizeable domestic workforce in strategic sectors. They have an inherent multiplier effect that politicians can leverage, for instance to create domestic jobs or increase wages. Furthermore, they have a large workforce that could shield them from regulation, making protectionism more likely if it would prove favourable for them. They often create so-called national brands – an extension of the national champions themselves. These brands are typically shielded against protectionist measures thanks to their diversified, loyal and less price-sensitive customer base, and their global reach. These brands also benefit from government lobbying, as Airbus did during French President Emmanuel Macron’s visit to China in January 2018, during which the Chinese government signalled an interest in buying 184 Airbus jets.

Each of these geopolitical challenges is different, and each reflects the distinctive interests, ambitions, and history of the country undertaking it. Yet there is growing cooperation between the countries that are challenging the regional pillars of the U.S.-led order. Russia and China have collaborated on issues such as energy, sales and development of military technology, opposition to additional U.S. military deployments on the Korean peninsula, and naval exercises from the South China Sea to the Baltic. In Syria, Iran provided the shock troops that helped keep Russia’s ally, Bashar al-Assad, in power, as Moscow provided the air power and the diplomatic cover. “Our cooperation can isolate America,” supreme leader Ali Khamenei told Putin in 2017. More broadly, what links these challenges together is their opposition to the constellation of power, norms, and relationships that the U.S.-led order entails, and in their propensity to use violence, coercion, and intimidation as means of making that opposition effective. Taken collectively, these challenges constitute a geopolitical sea change from the post-Cold War era. The revival of great-power competition entails higher international tensions than the world has known for decades, and the revival of arms races, security dilemmas, and other artifacts of a more dangerous past. It entails sharper conflicts over the international rules of the road on issues ranging from freedom of navigation to the illegitimacy of altering borders by force, and intensifying competitions over states that reside at the intersection of rival powers’ areas of interest. It requires confronting the prospect that rival powers could overturn the favorable regional balances that have underpinned the U.S.-led order for decades, and that they might construct rival spheres of influence from which America and the liberal ideas it has long promoted would be excluded. Finally, it necessitates recognizing that great-power rivalry could lead to great-power war, a prospect that seemed to have followed the Soviet empire onto the ash heap of history. Both Beijing and Moscow are, after all, optimizing their forces and exercising aggressively in preparation for potential conflicts with the United States and its allies; Russian doctrine explicitly emphasizes the limited use of nuclear weapons to achieve escalation dominance in a war with Washington. In Syria, U.S. and Russian forces even came into deadly contact in early 2018. American airpower decimated a contingent of government-sponsored Russian mercenaries that was attacking a base at which U.S. troops were present, an incident demonstrating the increasing boldness of Russian operations and the corresponding potential for escalation. The world has not yet returned to the epic clashes for global dominance that characterized the twentieth century, but it has returned to the historical norm of great-power struggle, with all the associated dangers. Those dangers may be even greater than most observers appreciate, because if today’s great-power competitions are still most intense at the regional level, who is to say where these competitions will end? By all appearances, Russia does not simply want to be a “regional power” (as Obama cuttingly described it) that dominates South Ossetia and Crimea.37 It aspires to the deep European and extra-regional impact that previous incarnations of the Russian state enjoyed. Why else would Putin boast about how far his troops can drive into Eastern Europe? Why else would Moscow be deploying military power into the Middle East? Why else would it be continuing to cultivate intelligence and military relationships in regions as remote as Latin America? Likewise, China is today focused primarily on securing its own geopolitical neighborhood, but its ambitions for tomorrow are clearly much bolder. Beijing probably does not envision itself fully overthrowing the international order, simply because it has profited far too much from the U.S.-anchored global economy. Yet China has nonetheless positioned itself for a global challenge to U.S. influence. Chinese military forces are deploying ever farther from China’s immediate periphery; Beijing has projected power into the Arctic and established bases and logistical points in the Indian Ocean and Horn of Africa. Popular Chinese movies depict Beijing replacing Washington as the dominant actor in sub-Saharan Africa—a fictional representation of a real-life effort long under way. The Belt and Road Initiative bespeaks an aspiration to link China to countries throughout Central Asia, the Middle East, and Europe; BRI, AIIB, and RCEP look like the beginning of an alternative institutional architecture to rival Washington’s. In 2017, Xi Jinping told the Nineteenth National Congress of the Chinese Communist Party that Beijing could now “take center stage in the world” and act as an alternative to U.S. leadership.38 These ambitions may or may not be realistic. But they demonstrate just how significantly the world’s leading authoritarian powers desire to shift the global environment over time. The revisionism we are seeing today may therefore be only the beginning. As China’s power continues to grow, or if it is successful in dominating the Western Pacific, it will surely move on to grander endeavors. If Russia reconsolidates control over the former Soviet space, it may seek to bring parts of the former Warsaw Pact to heel. Historically, this has been a recurring pattern of great-power behavior—interests expand with power, the appetite grows with the eating, risk-taking increases as early gambles are seen to pay off.39 This pattern is precisely why the revival of great-power competition is so concerning—because geopolitical revisionism by unsatisfied major powers has so often presaged intensifying international conflict, confrontation, and even war. The great-power behavior occurring today represents the warning light flashing on the dashboard. It tells us there may be still-greater traumas to come. The threats today are compelling and urgent, and there may someday come a time when the balance of power has shifted so markedly that the postwar international system cannot be sustained. Yet that moment of failure has not yet arrived, and so the goal of U.S. strategy should be not to hasten it by giving up prematurely, but to push it off as far into the future as possible. Rather than simply acquiescing in the decline of a world it spent generations building, America should aggressively bolster its defenses, with an eye to preserving and perhaps even selectively advancing its remarkable achievements.

“AI is too important not to regulate—and too important not to regulate well,” says Google. It is highly likely that Artificial Intelligence will be regulated. In many ways, it already is. The EU’s AI Act has extensive top-down prescriptive rules including prohibiting uses of AI that it says pose unacceptable risk; it’s in the final stages of passing. China has ruled that algorithms must be reviewed in advance by the state and “should adhere to the core socialist values”. The United States is taking its typical decentralized approach. The most likely outcome for the United States is a bottom-up patchwork quilt of executive branch actions. Unlike Europe, the United States is not likely to pass a broad national AI law over the next few years. Successful legislation is likely focused on less controversial and targeted measures like funding AI research and AI child safety. This likely disappoints proponents of strong national AI regulation. This outcome will be messier and have gaps, but no broad national law does not mean no regulation. We’re likely to see domain-specific agency actions especially in health care, financial services, housing, workforce, and child safety, plus multiple executive orders. This patchwork quilt of rules, if implemented well, could be grounded in the expertise of specific agencies and more tailored to innovation. The U.S. government will likely boost spending on AI and AI research, especially in defense and intelligence, and use its buying power to shape the market. AI trade friction with Europe is likely to emerge, and private companies will advance their own “responsible AI” initiatives and face a fragmented global AI regulatory landscape. Looming China competition will escalate a “don’t fall behind” debate. The Federal Trade Commission (especially) and Department of Justice likely fire warning shots to forestall AI concentration in big tech. There is a real but less likely chance that a key U.S. state (e.g., California) passes major AI legislation, or that a big AI-related disaster leads to a strong national response. This viewpoint proposes ten key parameters of successful AI regulatory design and the likely outcomes. AI raises many profound questions. Nearly all agree that AI promises both huge benefits for society and yet poses major risks. The challenge is getting the balance right between innovation and societal risks – which governments don't have a good record of achieving. The devil is in the details, making rules adaptable for a technology that is likely to change rapidly and be pervasive. The existential question of Artificial General Intelligence (AGI) is generating much debate. However, this is more distant and tough to forecast accurately. We focus here on issues like model safety, bias, transparency, privacy, security, trust, copyright, content regulation, education, and economic impacts (job loss, work force adjustment, productivity). For commercial actors, the key question is regulation’s impact on innovation, and does it address society’s concerns so that adoption is broad and sustainable. I. The AI Regulation Debate: A Wide Spectrum of Views. The AI regulation debate has become heated and politicized, reflecting the immense hopes and fears that we have invested in AI. Hysterical fear and paranoia. On one end, leading technologist and venture capitalist Marc Andreessen says “we have a full-blown moral panic about AI right now.” He views this as “hysterical fear” that is “irrational” and compares it to earlier introduction of new technologies that were similarly feared but were overblown and subsequently managed. He believes that AI is simply a computer program that is owned, controlled and used by people, and that AI does not have its own goals or its own wants. Moreover, he emphasizes, AI can be a force for tremendous good.

None of these approaches has received overwhelming support by study participants. With 56% of survey respondents echoing the current stance of the US Copyright Office, it appears that we are not yet at the proverbial crossroads requiring a change of direction in AI copyright policy. Divided opinion on the preparedness of the Copyright Office, however, signals the looming crisis at the intersection of artificial intelligence and intellectual property rights. VI. Conclusion This study has collected data from fifty-seven AI scientists, tech policy experts, and copyright scholars through a survey and questionnaire. The data confirms the significant role of intelligent computer programs in the American entertainment/creative industries. In addition, half of participants believe that the US copyright system is not adequately prepared for a future influx of AI-produced works. Respondents, however, fail to reach a resounding consensus on what changes should be implemented by the US Copyright Office. The divided nature of expert opinion and the limited data available to researchers studying intellectual property protection of AI works indicates the need for future research on the topic. This study has laid the groundwork for further investigation into the intersection of artificial intelligence and copyright protection. The survey and questionnaire have established a need for additional data and interdisciplinary research on a growing issue with the potential to significantly disrupt American economic and cultural development.

Designed to serve the crucial yet prosaic function of registering copyright claims, the United States Copyright Office (“Office”) increasingly asserts wide- ranging authority to interpret copyright law, advise courts and federal agencies, promulgate administrative and substantive rules, and broadly shape copyright policy. 1 Recent years have witnessed a flurry of Office rulemaking and policy studies addressing issues from bulk registration2 and the location of copyright notices 3 to resale royalties 4 and orphan works. 5 More controversially, the Office is in the midst of its seventh triennial rulemaking under § 1201 of the Digital Millennium Copyright Act (DMCA) 6 and has waded into, among other topics, copyright’s application to software-enabled consumer devices like cars, kitchen appliances, and medical devices. 7 Should security researchers be required to disclose the flaws they uncover to device makers and developers before going public? Should farmers be able to repair their tractors over the objections of manufacturers? These are the sorts of bizarrely incongruous questions that the Office confronts today. At the same time, the Office faces nagging constitutional questions about the scope of its authority and an often contentious relationship with the Library of Congress, of which it remains a subordinate office. Motivated in part by these concerns, some have called for greater independence for the Office. These proposals are closely linked to ongoing efforts to concentrate additional authority within the Office. To take a recent example, the Copyright Alternative in Small-Claims Enforcement (CASE) Act would establish a small claims tribunal within the Office that has the power to adjudicate infringement disputes.8 As the federal body responsible for administering copyright law, the Office offers considerable expertise on questions at the core of its competence. However, as a result of both Congressional mandates and the Office’s own ambitions, mission creep poses a problem. The Office regularly tackles complex economic and technological questions that extend well beyond both its historical mission and its most credible claims of subject matter expertise. Mission creep is a cause for concern for at least two reasons. First, through both the input it provides to Congress and its own direct regulatory action, the Office may advocate or adopt flawed policies. Second, the Office’s growing docket distracts it from its core task of developing and maintaining a registration system that serves the needs of copyright holders and the public. In some ways, the strains facing the Office are an unavoidable byproduct of an environment that is profoundly shaped by technological change and the expanding scope and significance of copyright law. Software copyright, technological protection measures, and digital distribution have altered the copyright landscape in dramatic ways. 9 Copyright law now regulates the use of software, both directly and indirectly, that is embedded in an astounding range of every day devices, even in the absence of traditional copyright infringement. 10 The need for experts to inform copyright policy in light of these developments is evident. Copyright law today implicates matters of security, privacy, and industrial competition that were entirely foreign to copyright policy debates just a few decades ago. But delegations and assertions of authority are insufficient to establish genuine expertise. If the Office intends to lead the policy conversation moving forward, it needs to build competence in those areas in which it is currently lacking. That process should begin with a frank assessment of the questions the Office is well-positioned to answer and those it is not.

Throughout this Part, we examined the potential pitfalls and frictions of applying copyright law principles to GAI, and how this proposed extension highlights a complex intersection of legal, ethical, and societal concerns. Such an expansion risks creating inequities by recognizing certain intellectual labor over others, reinforcing unfair labor hierarchies, and disregarding the invisibilized yet critical contributions of “ghost workers” in AI development. In invoking the concept of “copyright poisoning” my aim is to highlight the legal uncertainties and economic risks for industries engaging with GAI. Such poisoning leads to an unstable, volatile, and high-risk scenario, where AI advances are stifled by regulation that incentivizes litigation. Further still, major AI companies may also use copyright law as a competitive moat, hindering the entry of new players and perpetuating the major companies’ market dominance. The stringent application of copyright laws in this context could inadvertently hamper the development of beneficial AI applications across various sectors by imposing restrictive conditions on data usage. The potential overreach of copyright protections in the AI domain suggests a careful reconsideration of copyright laws is needed to ensure that the law fosters, rather than constrains, the dynamic landscape of digital innovation and promotes a robust public domain. CONCLUSION There are a plethora of ethical concerns about GAI that extend beyond the scope of copyright. With this Essay, I aim to reflect on how copyright might not be the adequate tool to fairly compensate artists and other data workers.

For the time being, transparency obligations should fall on the organizations that are developing and deploying AI models—not on consumers who are only end users of AI models. That may change if large numbers of end users themselves end up training and developing their own models, causing a rapid expansion in AI-associated resource consumption and byproduct production. However, the question about where transparency obligations fall must be revisited regularly as AI technologies continue to develop rapidly and increasingly resource-intensive queries by users become possible. Crucially, hypothetical future affordances of AI must not be factored into the answers to these sustainability questions. For example, if the goal of an AI initiative is to help an end user reduce their carbon emissions, then that hypothetical future reduction must not be factored into the organization’s assessment of the carbon emissions of this AI initiative this year. Policymakers should promote the monitoring and reporting of accurate information, rather than define “good” answers to these questions and penalize companies that do not meet those benchmarks. The EU’s Sustainable Finance Disclosure Regulation framework, with its emphasis on the power of transparency to shape and amplify market forces, can serve as a model for such an approach. If reported data were gathered in a single, open-access database (perhaps analogous to the European Single Access Point), then regulators, investors, technology companies, nonprofits, and the general public would be able to reward progress toward sustainability goals, over various time horizons, through a variety of mechanisms. It will be important to have external auditors to ensure the credibility of reported data, as they have done for sustainable finance.

To develop and leverage AI at the scale imagined by researchers, businesses, and governments, significant physical resources will be required for the design and deployment of the requisite computing hardware and software. While all AI approaches merit attention regarding their sustainability, generative AI is particularly resource-intensive: One such AI-powered chatbot is reportedly consuming the energy equivalent of 33,000 homes. (Note that while it is complicated to estimate such equivalences—given variations in operational timescales and details, home location, user numbers, etc.—various calculations have yielded estimated energy use equivalent to that of tens to hundreds of thousands of US homes.) In addition, new data centers are being designed and built with high demand and at a fast pace, new AI-critical hardware components are being designed and fabricated, and organizations large and small are experiencing urgency in setting their short-term tactics and long-term strategies for AI. Demands on data centers will only continue to grow as AI-powered applications spread through industries and around the world. For example, a recent International Energy Agency report projected an increase in data center energy consumption in 2026 equivalent to the energy consumption of Japan. Sustainability-focused regulation of AI, if deployed in a timely manner, can incentivize further improvements in the efficiency of data center operation and even the efficiency of software itself. Unfortunately, in the past, similar opportunities to promote the sustainable development of emerging technologies across industries have been missed. Failure to act during the rise of cryptocurrency mining has led to concerns today about the industry’s electricity and water use and to tension—internationally and domestically—around regulation and resource accessibility. For example, blockchain advocates filed a lawsuit against the US Department of Energy after the agency attempted to conduct an emergency survey of energy use by crypto miners, with the advocates arguing that it forced businesses to divulge sensitive information. More broadly, global digitization and its associated technologies have spurred crises in e-waste, supply-chain fragility, and human rights, to name a few. Early consideration and prioritization of these issues could have prevented harmful patterns from becoming embedded in today’s systems and processes. Crucially, the projected demands on data centers in the coming years due to the rise of AI—in terms of hardware, power, cooling, land and water use, and access to physical infrastructure and network bandwidth (a particular concern in growing urban areas)—are likely to far outstrip demands associated with other technologies. The potential cumulative impacts of the AI revolution, including resource consumption and byproduct production, underscore the urgency of acting today.

Sustainability, dignity and solidarity are significantly underrepresented compared to other ethical dimensions, which suggests that these issues might be currently flying under the radar of the mainstream AI ethics debate. This underrepresentation is particularly problematic in light of recent evidence that AI requires massive computational resources, which, in turn, require high energy consumption129. The environmental impact of AI, however, involves not only the negative effects of high-footprint digital infrastructures, but also the possibility of harnessing AI for the benefit of ecosystems and the entire biosphere. As the humanitarian cost of anthropogenic climate change is rapidly increasing130, the principles of sustainability and solidarity appear strictly intertwined. In addition, the ethical principle of solidarity is referenced even more rarely, typically in association with the development of inclusive strategies for the prevention of job losses and unfair sharing of burdens. Little attention is devoted to promoting solidarity through the emerging possibility of using AI expertise for solving humanitarian challenges, a mission that is currently being pursued, among others, by companies such as Microsoft131 and intergovernmental organizations such as the United Nations Office for Project Services (UNOPS)132 or the World Health Organization (WHO). A better appraisal of currently underrepresented ethical principles is likely to result in a more inclusive AI ethics landscape.

As much of the potential legislation around AI requires reviews of models, groups have begun offering certification programs to AI companies to show how their work doesn’t violate copyright. Bloomberg reported Fairly Trained — founded by former Stability AI vice president for audio Ed Newton-Rex — adds a label to companies that prove they asked for permission to use copyrighted training data. Newton-Rex started Fairly Trained after he quit Stability AI in November, citing generative AI was “exploiting creators.” Its first accreditation, which it calls the Licensed Model certification, will be awarded to companies that license protected data to train its models. Fairly Trained claims it will not issue the certification to developers that rely on the fair use argument to train models. Fairly Trained said in a blog post that it has already certified nine generative AI companies that work in image, music, and voice generation. These include Beatoven.ai, Boomy, BRIA.ai, Endel, LifeScore, Rightsify, SOMMS.AI, Soundful, and Tuney. Training AI models with copyrighted data has been a fraught issue in generative AI, especially after artists and authors sued several AI companies for copyright infringement. The New York Times also filed a lawsuit against OpenAI and Microsoft for violating its copyright when training its GPT models. Some proposed legislation currently making its way through Congress would require AI companies to disclose where they got their training data. The information will let copyright holders know if their work was used without permission.

US: Efforts in the Three Branches and the Marketplace to Resolve the AI and Copyright Dilemma Executive Branch and Copyright Office Efforts on AI and Copyright Here in the United States, in August 2023, the US Copyright Office published a request for comments on the question of how the office should address governance challenges posed by the emergence of generative AI. Among other issues, the Copyright Office requested comment on “whether legislative or regulatory steps” are needed to address “the use of copyrighted works to train AI models.” The comment period for this inquiry closed on December 6, 2023. The August 2023 inquiry, along with previous efforts by the Copyright Office, are intended to provide the material for future legislative reforms. Policy proposals on AI and copyright made on the basis of the Copyright Office’s inquiries may come as early as summer 2024. President Joe Biden’s October 2023 “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence” directs the Copyright Office to “issue recommendations to the President on potential executive actions relating to copyright and AI… including the scope of protection for works produced using AI and the treatment of copyrighted works in AI training.” These recommendations are to be issued to the President by the end of July 2024 at the latest. Congressional Proposals on AI and Copyright In parallel with the executive branch and the Copyright Office, lawmakers in Congress are wrestling with the copyright issues posed by AI. On December 22, 2023, Representatives Don Beyer (D-VA-8) and Anna Eshoo (D-CA-16), respectively the Vice Chair and Co-Chair of the Congressional AI Caucus, introduced the AI Foundation Model Transparency Act. Under the act, developers of certain AI models may be required to disclose the sources of data used to train their models, including “information necessary to assist copyright owners or data license holders with enforcing their copyright or data license protections...” As members and committees propose piecemeal AI bills, Congressional leadership has been working towards the goal of comprehensive AI legislation. From September to December 2023, a bipartisan group of senators led by Majority Leader Chuck Schumer (D-NY) hosted a series of senate-wide “AI Insight Forums,” or closed-door meetings with a variety of stakeholders on pertinent governance issues raised by AI. The purpose of these forums is to develop expertise within the Senate needed to produce effective bi-partisan AI legislation within “months.” On November 29, 2023, Senator Schumer hosted a forum on “Transparency, Explainability, Intellectual Property, and Copyright.” In his opening statement for the forum, the senator asserted that “there is a role for Congress to play to promote transparency in AI systems and protect the rights of creators.” Will the Courts Decide the Issue Under Current Law? On December 27, 2023, the New York Times sued leading generative AI companies for “unlawful use of The Times’s work to create artificial intelligence products that compete with” the publication. The complaint alleges that the defendants used “Times content without permission to develop their models and tools.” According to the complaint, after a months-long attempt to reach a negotiated agreement stalled, the publication filed suit. The New York Times argues that the defendants’ use of “The Times’s content without payment to create products that… closely mimic the inputs used to train them” is not protected by fair use and therefore constitutes copyright infringement. Given the high profile of both the plaintiff and defendants, some experts have predicted that this case, if tried, will eventually be appealed to the Supreme Court. Regardless of whether this ends up being the case, the outcome of this suit could have significant implications for not just the news industry, but all creators whose content has been used to train AI models. In the absence of regulation from the executive or legislative branch, a decisive ruling on this or a related case may set industry standards regarding the use of copyrighted works as training data for AI models—or be the catalyst for codifying a set of standards. Self-Regulation as the Solution? Aside from government agencies, Congress, and the courts, the ultimate solution to the issues posed by the use of copyrighted works as AI training data may come from the marketplace. Prior to filing their suit, the New York Times engaged with generative AI companies in months of negotiations over the use of its content. Had these negotiations succeeded, the resulting arrangement could have been used as a model for other publishers and content creators. As the New York Times suit progresses, there is nothing preventing marketplace agreements (including from the litigants). The emergence of AI is moving at warp speed. Regulatory, legislative, and litigation-based solutions all move at a slower pace. In all likelihood, the longer it takes for regulators to create and implement public policy solutions to this issue, the more likely it is that creators and AI developers will come to their own arrangements.

As the president and policymakers consider rules to govern AI, they face intractable challenges. Trying to regulate every expression of AI will be an elusive exercise; the technology is far too ingrained already in our lives to even think it possible. Equally important, government regulation, however well-intentioned, will always be several steps behind even in its best iteration. All things considered, self-regulation is the best policy course for AI regulation. Here is why: Industry expertise AI is vast, complex and rife with nuances. Technology experts immersed in its intricacies are best suited to draft and enforce guidelines. Even the most expert federal regulatory agencies, while motivated, are sure to lack the depth of understanding necessary to effectively oversee AI innovations. Those within the AI industry possess the hands-on experience and technical familiarity to establish guidelines that are both practical and effective. Agility and responsiveness The pace of AI advancement is dynamic and staggering. By the time traditional regulatory frameworks catch up with the latest development, several new ones will have emerged. Self-regulation offers a flexibility that can adapt swiftly to the ever-changing AI landscape. This agility ensures that guidelines remain relevant and fosters responsible innovation without stifling it. Cost-effectiveness The bureaucratic nature of government-led regulations often entails prolonged processes and significant financial expenditures. In contrast, self-regulation, led by industry stakeholders, can be more streamlined and cost-effective, reducing the financial burden on both the industry and taxpayers. Preempting stricter oversight In the absence of self-regulation, government might impose even stricter, potentially stifling regulations. By adopting a proactive stance, the AI industry can set standards that are both high and realistic, preempting the need for burdensome regimes. Building trust through transparency One of the critiques of AI is the perceived “black box” nature of algorithms, which can lead to mistrust among regulators and the public. By self-imposing transparency and ethical standards, the AI industry can build credibility and trust. To be fair, there are downsides to self-regulation that augur for government regulation. Proponents of federal regulation argue that the Federal Trade Commission (FTC) and Department of Justice would be more inclined than industry to protect individual rights, monitor and prevent abuses and strictly enforce the rules. With concerns over AI’s influence on privacy, federal regulations can provide a standardized framework to ensure that personal data is not misused. In areas like autonomous vehicles and healthcare AI, regulations can ensure the development and deployment of safe, reliable AI technologies. Without proper oversight, AI systems can inadvertently perpetuate or even amplify biases present in the data they are trained on. Regulation can ensure systems are developed and deployed in a manner that mitigates these biases. Regulations can require companies to make their AI systems more understandable and transparent, which can help in building public trust. By setting rigorous standards, the U.S. might position itself as a producer of high-quality, reliable and trustworthy AI solutions, which could become a selling point on the international stage. Just as the FDA ensures the safety of drugs and the FAA regulates aviation safety, AI regulation can be a way to protect consumers from potentially harmful or misleading AI-driven products and services. Regulations can also drive companies to adopt ethical AI practices, ensuring technology benefits humanity as a whole. But enforcing federal AI regulations can be challenging, given the complexity of AI itself. The risk of governmental overreach is as real as the infringement of individual freedom. As AI companies often operate globally, navigating a patchwork of different regulatory regimes across countries can be complex and costly. We must also be concerned about competition; if the U.S. over-regulates, it may fall behind in the global AI race. In addition, the economic impact of over-regulation might drive AI companies to relocate to more lenient jurisdictions, potentially leading to job losses and a downturn in economic contributions from a rapidly growing sector. Finally, ill-advised federal regulation has direct economic implications for small- and medium-sized businesses. Complying with regulations often comes with costs. While large corporations might easily absorb these, small businesses and startups could struggle, potentially hampering innovation at grassroots levels. To preserve and advance the many benefits of artificial intelligence, U.S. policymakers should encourage the industry to develop a robust and prudent self-regulatory regime that fosters accountability, ethical responsibility, innovation, and security. As AI continues to permeate our everyday lives, there is no doubt that regulation is necessary. But its direction should come from industry, not government, and be guided by a set of industry-derived principles and best practices that are fair, transparent and enforceable.

How does interdependence affect war, the most intense form of conflict? Table 2 gives the empirical results. The rarity of wars makes any analysis of their causes quite difficult, for variations in interdependence will seldom result in the occurrence of war. As in the case of MIDs, the log-likelihood ratio tests for each model suggest that the inclusion of the various measures of interdependence and the control variables improves our understanding of the factors affecting the occurrence of war over that obtained from the null model. However, the individual interdependence variables, alone, are not statistically significant. This is not the case with contiguity and relative capabilities, which are both statistically significant. Again, we see that contiguous dyads are more conflict-prone and that dyads composed of states with unequal power are more pacific than those with highly equal power. Surprisingly, no evidence is provided to support the commonly held proposition that democratic states are less likely to engage in wars with other democratic states.¶ The evidence from the pre-WWII period provides support for those arguing that economic factors have little, if any, influence on affecting leaders’ decisions to engage in war, but many of the control variables are also statistically insignificant. These results should be interpreted with caution, since the sample does not contain a sufficient number wars to allow us to capture great variations across different types of relationships. Many observations of war are excluded from the sample by virtue of not having the corresponding explanatory measures. A variable would have to have an extremely strong influence on conflict—as does contiguity—to find significant results. ¶ 7. Conclusions This study provides little empirical support for the liberal proposition that trade provides a path to interstate peace. Even after controlling for the influence of contiguity, joint democracy, alliance ties, and relative capabilities, the evidence suggests that in most instances trade fails to deter conflict. Instead, extensive economic interdependence increases the likelihood that dyads engage in militarized dispute; however, it appears to have little influence on the incidence of war.

Both parties protectionist now

Can Trump enact tariffs on his own? Though it would appear to contradict the Origination Clause of the Constitution, Congress has delegated that authority in enough pieces of legislation that Trump could probably raise import duties unilaterally. But what would be the practical effect? Hard-core free traders paint a picture of cataclysm. Tariffs will launch trade wars, increase prices, and destroy the economy. This is all hard-wired into the pro-globalization worldview. Thomas Friedman once famously admitted that he wrote a column supporting a free-trade agreement with Central America without knowing a thing about it: “I just knew two words: free trade,” he told an audience. Presumably the opposite is true for Friedman: He sees one word, “tariff,” and immediately screams in horror. Oddly, many of those same proponents of free trade favor a policy that looks very much like a tariff. The Republican corporate-tax revamp includes something called a border-adjustment tax, which would impose a 20 percent tax on imports while eliminating a tax on exports. Like with tariffs, the goal appears to be to encourage domestic production. In fact, the tax would be much higher than the 5-10 percent tariff being floated. (It also might be illegal under the current global trade regime.) Supporters of border adjustment, particularly economists, argue that it will end up trade neutral, because the exchange rate will fluctuate in response to the tax. In other words, though the tax would make American-made goods more attractive, the value of the dollar would increase, leveling that out. Few of these economists seem to carry over the same analysis to the effects of a tariff. I don’t understand why. There’s no reason to doubt the fact that, if Trump imposed an across-the-board tariff, the dollar would strengthen, thus nullifying the desired effect. Indeed, before Trump has even taken office, the dollar has risen to a 14-year high, in anticipation of a more protectionist stance. Incidentally, for all the one-off announcements by Trump (however factually challenged) about hundreds of jobs he has allegedly rescued here or there, this one development—the rise in the dollar—has likely caused the loss of hundreds of thousands of manufacturing jobs, under standard economic theory. Looked at this way, higher tariffs wouldn’t cause a recession (as Paul Krugman has acknowledged), but would be somewhat pointless, with currency exchanges shifting to account for any changes. Trade wars might temporarily reduce efficiency, as domestic supply chains would have to be rebuilt, but they’re unlikely to radically alter the balance of trade on their own. There are other variables here. Importers and exporters who have lived in a world of floating exchange rates for decades may be fairly nimble in adjusting to them. On the downside, Krugman explains that raising tariffs could inhibit capital flows, meaning that investors will place less money into US markets. You can see how that might reduce economic growth. But Jeff Spross points out that America currently has a problem with too much foreign money flowing in; reducing the flow could arguably make the economy more stable. Trump could also seek to prevent unlawful currency manipulation (not necessarily from China, but from other Asian nations) that artificially disadvantages US manufacturing. The real unknown here is what Trump would do with all that tariff revenue. The border adjustment tax at 20 percent is assumed to bring in $1 trillion over the 10-year budget window. So a tariff of even one-quarter or one-half that size would draw significant funds. What’s the plan for it? Would it get plowed into job-creating investments? Tax cuts for the wealthy? That’s a significant variable as well. We do know that the same pundits who confidently predicted that globalization would be a win-win policy for America repeatedly got it wrong. Those on the losing side saw their jobs shipped out and factories closed down, and weren’t given the kind of assistance needed to offset the disruption. So it’s worth being a little skeptical of the warnings coming from the same corners now. I don’t have a ton of faith in the Trump team to necessarily make their trade agenda work (especially as corporate interests will seek to co-opt the redesigned policies in ways even friendlier to their bottom line). And I think there are smarter ways to balance our trade deficit than a tariff strategy which will just run up against currency exchange rates. But the hysteria accompanying these tariffs (which wasn’t at all present when President Obama imposed his own tariffs on Chinese tires and steel) seems far beyond what little we can assume about the actual results of such a strategy.

Yet this "mother of all assumptions" is simply asserted and rarely examined. The obvious question to ask is this: did prior acts of nuclear proliferation have the same fearsome consequences that Iran hawks now forecast? The answer is no. In fact, the spread of nuclear weapons has had remarkably little impact on the basic nature of world politics and the ranking of major powers. The main effect of the nuclear revolution has been to induce greater caution in the behavior of both those who possessed the bomb and anyone who had to deal with a nuclear-armed adversary. Proliferation has not transformed weak states into influential global actors, has not given nuclear-armed states the ability to blackmail their neighbors or force them to kowtow, and it has not triggered far-reaching regional arms races. In short, fears that an Iranian bomb would transform regional or global politics have been greatly exaggerated; one might even say that they are just a lot of hooey. Consider the historical record. Did the world turn on its axis when the mighty Soviet Union tested its first bomb in 1949? Although alarmist documents like NSC-68 warned of a vast increase in Soviet influence and aggressiveness, Soviet nuclear development simply reinforcedthe caution that both superpowers were already displaying towards each other. The United States already saw the USSR as an enemy, and the basic principles of containment were already in place. NATO was being formed before the Soviet test and Soviet dominance of Eastern Europe was already a fait accompli. Having sole possession of the bomb hadn't enabled Truman to simply dictate to Stalin, and getting the bomb didn't enable Stalin or his successors to blackmail any of their neighbors or key U.S. allies. It certainly didn't lead any countries to "reorient their political alignment toward Moscow." Nikita Khrushchev's subsequent missile rattling merely strengthened the cohesion of NATO and other U.S.-led alliances, and we now know that much of his bluster was intended to conceal Soviet strategic inferiority. Having a large nuclear arsenal didn't stop the anti-commnist uprisings in East Germany, Hungary, Czechoslovakia, or Poland, and didn't allow the Soviet Union to win in Afghanistan. Nor did it prevent the USSR from eventually collapsing entirely. Did British and French acquisition of nuclear weapons slow their decline as great powers? Not in the slightest. Having the force de frappe may have made De Gaulle feel better about French prestige and having their own deterrent made both states less dependent on America's security umbrella, but it didn't give either state a louder voice in world affairs or win them new influence anywhere. And you might recall that Britain couldn't get Argentina to give back the Falklands by issuing nuclear threats -- even though Argentina had no bomb of its own and no nuclear guarantee -- they had to go retake the islands with conventional forces.  Did China's detonation of a bomb in 1964 suddenly make them a superpower? Hardly. China remained a minor actor on the world stage until it adopted market principles, and its rising global influence is due to three decades of economic growth, not a pile of nukes. And by the way, did getting a bomb enable Mao Zedong--a cruel megalomaniac who launched the disastrous Great Leap Forward in 1957 and the destructive Cultural Revolution in the 1960s -- to start threatening and blackmailing his neighbors? Nope. In fact, China's foreign policy behavior after 1964 was generally quite restrained. What about Israel? Does Israel's nuclear arsenal allow it to coerce its neighbors or impose its will on Hezbollah or the Palestinians? No. Israel uses its conventional military superiority to try to do these things, not its nuclear arsenal. Indeed, Israel's bomb didn't even prevent Egypt and Syria from attacking it in October 1973, although it did help convince them to limit their aims to regaining the territory they had lost in 1967. It is also worth noting that Israel's nuclear program did not trigger a rapid arms race either. Although states like Iraq and Libya did establish their own WMD programs after Israel got the bomb, none of their nuclear efforts moved very rapidly or made it across the finish line. But wait, there's more. The white government in South Africa eventually produced a handful of bombs, but nobody noticed and apartheid ended anyway. Then the new government gave up its nuclear arsenal to much acclaim. If anything, South Africa was more secure without an arsenal than it was before. What about India and Pakistan? India's "peaceful nuclear explosion" in 1974 didn't turn it into a global superpower, and its only real effect was to spur Pakistan -- which was already an avowed rival -- to get one too. And it's worth noting that there hasn't been a large-scale war between the two countries since, despite considerable grievances on both sides and occasional skirmishes and other provocations. Finally, North Korea is as annoying and weird as it has always been, but getting nuclear weapons didn't transform it from an economic basket case into a mighty regional power and didn't make it more inclined to misbehave. In fact, what is most remarkable about North Korea's nuclear program is how little impact it has had on its neighbors. States like Japan and South Korea could go nuclear very quickly if they wanted to, but neither has done so in the six years since North Korea's first nuclear test.  In short, both theory and history teach us that getting a nuclear weapon has less impact on a country's power and influence than many believe, and the slow spread of nuclear weapons has only modest effects on global and regional politics. Nuclear weapons are good for deterring direct attacks on one's homeland, and they induce greater caution in the minds of national leaders of all kinds. What they don't do is turn weak states into great powers, they are useless as tools of blackmail, and they cost a lot of money. They also lead other states to worry more about one's intentions and to band together for self-protection. For these reasons, most potential nuclear states have concluded that getting the bomb isn't worth it. But a few states-and usually those who are worried about being attacked-decide to go ahead. The good news is that when they do, it has remarkably little impact on world affairs. 

The proliferation of nuclear weapons has been far slower than predicted, and this seems to be because the weapons do not generally convey much advantage to their possessor. Dozens of technologically capable countries have considered obtaining nuclear arsenals, but very few have done so. A key reason for this is that the possession of such expensive armaments actually conveys in almost all cases rather little advantage to the possessor. In the main, they are difficult to obtain, militarily useless, a spectacular waste of time, money, and scientific talent, distasteful, and likely to rile the neighbors. Moreover, the weapons have also been exceedingly difficult to obtain for administratively dysfunctional countries.9

Moreover, the consequences of what little proliferation has taken place have been substantially benign. Those who have acquired the weapons have “used” them simply to stoke their egos or to deter real or imagined threats. For the most part, they have quietly kept the weapons in storage and haven’t even found much benefit in rattling them from time to time.

The value of democracy has been taken for granted until recently, but this assumption seems to be under threat now more than ever before. As was explained in Chapter 1, democracy’s claim to be valuable does not rest on just one particular merit, and scholars tend to distinguish three different types of values (Sen 1999). This book focused on the instrumental value of democracy (and hence not on the intrinsic and constructive value), and investigated the value of democracy for peace (Chapters 3 and 4), control of corruption (Chapter 5) and economic development (Chapter 6). This study was based on a search of an enormous academic database for certain keywords,6 then pruned the thousands of articles down to a few hundred articles (see Appendix) which statistically analysed the connection between the democracy and the four expected outcomes. The first finding is that a reverse wave away from democracy has not happened (see Chapter 2). Not yet, at least. Democracy is not doing worse than before, at least not in comparative perspective. While it is true that there is a dramatic decline in democracy in some countries,7 a general trend downwards cannot yet be detected. It would be better to talk about ‘stagnation’, as not many dictatorships have democratized recently, while democracies have not yet collapsed. Another finding is that the instrumental value of democracy is very questionable. The feld has been deeply polarized between researchers who endorse a link between democracy and positive outcomes, and those who reject this optimistic idea and instead emphasize the negative effects of democracy. There has been ‘no consensus’ in the quantitative literature on whether democracy has instrumental value which leads some beneficial general outcomes. Some scholars claim there is a consensus, but they only do so by ignoring a huge amount of literature which rejects their own point of view. After undertaking a large-scale analysis of carefully selected articles published on the topic (see Appendix), this book can conclude that the connections between democracy and expected benefits are not as strong as they seem. Hence, we should not overstate the links between the phenomena. The overall evidence is weak. Take the expected impact of democracy on peace for example. As Chapter 3 showed, the study of democracy and interstate war has been a flourishing theme in political science, particularly since the 1970s. However, there are four reasons why democracy does not cause peace between countries, and why the empirical support for the popular idea of democratic peace is quite weak. Most statistical studies have not found a strong correlation between democracy and interstate war at the dyadic level. They show that there are other—more powerful—explanations for war and peace, and even that the impact of democracy is a spurious one (caveat 1). Moreover, the theoretical foundation of the democratic peace hypothesis is weak, and the causal mechanisms are unclear (caveat 2). In addition, democracies are not necessarily more peaceful in general, and the evidence for the democratic peace hypothesis at the monadic level is inconclusive (caveat 3). Finally, the process of democratization is dangerous. Living in a democratizing country means living in a less peaceful country (caveat 4). With regard to peace between countries, we cannot defend the idea that democracy has instrumental value. Can the (instrumental) value of democracy be found in the prevention of civil war? Or is the evidence for the opposite idea more convincing, and does democracy have a ‘dark side’ which makes civil war more likely? The findings are confusing, which is exacerbated by the fact that different aspects of civil war (prevalence, onset, duration and severity) are mixed up in some civil war studies. Moreover, defining civil war is a delicate, politically sensitive issue. Determining whether there is a civil war in a particular country is incredibly difficult, while measurements suffer from many weaknesses (caveat 1). Moreover, there is no linear link: civil wars are just as unlikely in democracies as in dictatorships (caveat 2). Civil war is most likely in times of political change. Democratization is a very unpredictable, dangerous process, increasing the chance of civil war significantly. Hybrid systems are at risk as well: the chance of civil war is much higher compared to other political systems (caveat 3). More specifically, both the strength and type of political institutions matter when explaining civil war. However, the type of political system (e.g. democracy or dictatorship) is not the decisive factor at all (caveat 4). Finally, democracy has only limited explanatory power (caveat 5). Economic factors are far more significant than political factors (such as having a democratic system) when explaining the onset, duration and severity of civil war. To prevent civil war, it would make more sense to make poorer countries richer, instead of promoting democracy. Helping countries to democratize would even be a very dangerous idea, as countries with changing levels of democracy are most vulnerable, making civil wars most likely. It is true that there is evidence that the chance of civil war decreases when the extent of democracy increases considerably. The problem however is that most countries do not go through big political changes but through small changes instead; those small steps—away or towards more democracy—are dangerous. Not only is the onset of civil war likely under such circumstances, but civil wars also tend to be longer, and the conflict is more cruel leading to more victims, destruction and killings (see Chapter 4). A more encouraging story can be told around the value for democracy to control corruption in a country (see Chapter 5). Fighting corruption has been high on the agenda of international organizations such as the World Bank and the IMF. Moreover, the theme of corruption has been studied thoroughly in many different academic disciplines—mainly in economics, but also in sociology, political science and law. Democracy has often been suggested as one of the remedies when fighting against high levels of continuous corruption. So far, the statistical evidence has strongly supported this idea. As Chapter 5 showed, dozens of studies with broad quantitative, cross-national and comparative research have found statistically significant associations between (less) democracy and (more) corruption. However, there are vast problems around conceptualization (caveat 1) and measurement (caveat 2) of ‘corruption’. Another caveat is that democratizing countries are the poorest performers with regard to controlling corruption (caveat 3). Moreover, it is not democracy in general, but particular political institutions which have an impact on the control of corruption; and a free press also helps a lot in order to limit corruptive practices in a country (caveat 4). In addition, democracies seem to be less affected by corruption than dictatorships, but at the same time, there is clear evidence that economic factors have more explanatory power (caveat 5). In conclusion, more democracy means less corruption, but we need to be modest (as other factors matter more) and cautious (as there are many caveats). The perceived impact of democracy on development has been highly contested as well (see Chapter 6). Some scholars argue that democratic systems have a positive impact, while others argue that high levels of democracy actually reduce the levels of economic growth and development. Particularly since the 1990s, statistical studies have focused on this debate, and the empirical evidence is clear: there is no direct impact of democracy on development. Hence, both approaches cannot be supported (see caveat 1). The indirect impact via other factors is also questionable (caveat 2). Moreover, there is too much variation in levels of economic growth and development among the dictatorial systems, and there are huge regional differences (caveat 3). Adopting a one-size-fits-all approach would not be wise at all. In addition, in order to increase development, it would be better to focus on alternative factors such as improving institutional quality and good governance (caveat 4). There is not sufficient evidence to state that democracy has instrumental value, at least not with regard to economic growth. However, future research needs to include broader concepts and measurements of development in their models, as so far studies have mainly focused on explaining cross-national differences in growth of GDP (caveat 5). Overall, the instrumental value of democracy is—at best—tentative, or—if being less mild—simply non-existent. Democracy is not necessarily better than any alternative form of government. With regard to many of the expected benefits—such as less war, less corruption and more economic development—democracy does deliver, but so do nondemocratic systems. High or low levels of democracy do not make a distinctive difference. Mid-range democracy levels do matter though. Hybrid systems can be associated with many negative outcomes, while this is also the case for democratizing countries. Moreover, other explanations—typically certain favourable economic factors in a country—are much more powerful to explain the expected benefits, at least compared to the single fact that a country is a democracy or not. The impact of democracy fades away in the powerful shadows of the economic factors.8

In “Subjective and Objective Measures of Democratic Backsliding,” the political scientists Andrew Little of the University of California at Berkeley and Anne Meng of the University of Virginia find that the post-Cold War expansion of democracy in nations around the world has, on average, held. “Recent studies that find evidence of global backsliding,” they observe, “rely heavily if not entirely on subjective indicators” — such as asking experts to rate, based on their own judgment, the extent to which an election was “free and fair.”

Little and Meng instead measure democracy based on quantifiable factors less susceptible to individual bias. They find, for example, no overall decline in electoral competition. Across the world, “The rate of leader and ruling party turnover has remained fairly constant since the late 1990s,” the paper says. “If anything, the vote shares of incumbent leaders in executive elections and incumbent parties in legislatures have decreased in recent years.”

Resolution of this controversy is fundamental to the study and practice of international relations. The observation of democratic peace is ‘‘the closest thing we have to an empirical law’’ in the study of global politics (Levy, 1988: 662), and carries the profound implication that the spread of democracy will end war. New economic norms theory, on the other hand, yields the contrary implication that universal democracy will not end war. Instead, it is market-oriented development that creates a culture of contracting, and this culture legitimates democracy within nations and causes peace among them. The policy implications could hardly be more divergent: to end war (and support democracy), the contractualist democracies should promote the economies of nations at risk (Krieger and Meierrieks, 2015; Meierrieks, 2012; Mousseau, 2000, 2009, 2012a, 2013; Nieman, 2015).

In the literature are five factual claims for why we should continue to believe that democracy causes peace: (1) an assertion that in three of the five studies that overturned the democratic peace (Mousseau, 2013; Mousseau et al., 2013a, b), the insignificance of democracy controlling for contractualist economy is due to the treatment of missing data for contractualist economy (Dafoe et al., 2013, henceforth DOR); (2) a claim of error in the measure for conflict (DOR) that appears in one of the five studies that overturned the democratic peace (Mousseau, 2013); (3) an alleged misinterpretation of an interaction term that appears in one of the five studies (Mousseau, 2009) that overturned the democratic peace, along with in inference of democratic causality from an interaction of democracy with contractualist economy (Dafoe and Russett, 2013; DOR); (4) a claim of reverse causality, of democracy causing contractualist economy (Ray, 2013); and (5) a report of multiple regressions with most said to show democratic significance after controlling for contractualist economy (DOR).

It’s easy to construct climate disasters. You just find a current, disconcerting trend and project it into the future, while ignoring everything humanity could do to adapt. For instance, one widely reported study found that heat waves could kill thousands more Americans by the end of the century if global warming continues apace—but only if you assume people won’t use more air conditioning. Yes, the climate is likely to change, but so is human behavior in response.

Adaptation doesn’t make the cost of global warming go away entirely, but it does reduce it dramatically. Higher temperatures will shrink harvests if farmers keep growing the same crops, but they’re likely to adapt by growing other varieties or different plants altogether. Corn production in North America has shifted away from the Southeast toward the Upper Midwest, where farmers take advantage of longer growing seasons and less-frequent extreme heat. When sea levels rise, governments build defenses—like the levees, flood walls and drainage systems that protected New Orleans from much of Hurricane Ida’s ferocity this year.