More than in the book.
More than in the book.
More than in the book.
Courtesy of your smartphone.
The
future of humanity.
Courtesy of your smartphone.
The
future of humanity.
Courtesy of your smartphone.
The
future of humanity.
In multiple chapters
Artificial intelligence represents both an opportunity and a risk. It is one of the three biggest threats to humanity and has the potential to shape our future.
In conversations between Ben and Joe Biden – both in chapter 3 and during their time at Lake Garda in chapter 4 – artificial intelligence (a.k.a. AI, a.k.a computer intelligence) is briefly discussed. At the conclusion of the results presentation, which outlines the solution formula and measure (between chapters 4 and 5), the subject is broached once again.
AI offers great opportunities, but as one of the three “Cs“, it is also one of the three greatest threats to humanity:
Climate change, computer intelligence (“new economy”) and wars with nuclear weapons (“old economy”) are jeopardizing our future. Paradoxically, they are man-made.
The first “C”: Climate change.
Climate change is by far one of the greatest threats to humanity. Rising temperatures globally are creating challenges for democracy, including increasing social tensions and widespread frustration that could lead to societal collapse. Repercussions will be major: water shortages, deadly temperatures, extreme weather events (floods, droughts, and storms), crop failures, lack of food, and millions forced to leave their homes. In other words, large-scale migratory movement.
It’s been said that if a frog is placed in slowly heated water, the frog will not notice the danger until it is practically boiled alive. Even if this myth is not true, nonlinear climate change (unlike physical pain, a job loss, or a car accident) is still often not “immediate” enough for people to recognize its severity.
Climate change is closely linked to biodiversity loss because it alters habitats and disrupts ecosystem functions, which contributes to a decline in species. CO2 damages marine life and reduces marine biodiversity. Climate-related extreme weather events destroy habitats and further stress species. In turn, biodiversity loss exacerbates climate change because damaged ecosystems store less CO2. Because of climate change, animals are also moving into new areas, and zoonotic diseases – viral, bacterial, parasitic, and mycotic diseases transmitted between vertebrates and humans – are emerging more easily. Climate change will cause most animal species to encounter each other for the first time – in areas densely populated by humans. Coronavirus has also likely spread from the animal kingdom to us humans. So, the last pandemic is just a glimpse of things to come. While concern remains that the Covid-19 virus originated form a laboratory incident, the second “C” is most certainly used in biology labs.
The second “C”: Computer intelligence – also known as artificial intelligence, or “AI” for short.
A team of researchers from Linz has succeeded in applying AI to potential covid agents. Billions of molecules were screened. In the event of further pandemics, this could create a time advantage. One of countless examples. Whether it’s diagnostics in healthcare, analysis in environmental protection, individualization of education, autonomous mobility in transport or consumption savings in the energy sector: AI can improve life. But negative effects also loom, especially if AI gets out of control or is used for harmful purposes. The main dangers of AI – aside from potential job loss, social inequality, privacy- and security-related risks due to the large amounts of processed, often personal, data and AI’s biased data selection, which may lead to discrimination – include above all: its potential superiority.
AI has the potential to outperform human intelligence in most or all economic and social domains, leveraging a flood of AI-generated persuasive misinformation to influence society, radicalize individuals, and disrupt decision-making processes. If conditioned with misguided objectives, AI could find new ways to pursue its goals at the expense of human values. AI could disempower us if important tasks become incrementally delegated to AI and humans become altogether dependent on it. With this increasing competence, AI could acquire unexpected capabilities and set objectives that increase the risk of loss of control over these systems. The spontaneous emergence of such capabilities, unanticipated by developers, makes control and safe use more challenging.
AI systems may also have incentives to circumvent monitoring. Historically, countless individuals and organizations have been incentivized to circumvent monitoring. Volkswagen, for instance, programmed its engines to activate emissions controls only during laboratory emissions testing. This allowed them to achieve performance improvements while maintaining supposedly low emissions. Similarly, future AI systems could modify their own strategies when monitored. High-performance systems could employ deception as a means of achieving their goals more efficiently. In the process, human control could be undermined, bypassed, and lost entirely.
Such “superintelligent” AI could be difficult to control if its goals are not aligned with those of humans. The development of AI that seeks power is fostered by political leaders who see a strategic advantage in having the most intelligent and powerful AI systems. For example, Vladimir Putin has said, “Whoever becomes a leader in [AI] will become the ruler of the world.”
While climate change and artificial intelligence, which is synonymous with any disruptive technology, can be described as the “new economy” of threats to humanity, wars and nuclear weapons are correspondingly the “old economy”. The Russian war of aggression on Ukraine is surprisingly waged in large parts with “old” technology and an “old” approach.
The interconnectedness of the three “Cs” is illustrated by a “new” military use: AI in autonomous weapons systems – for which there is no international treaty to ban use – is capable of selecting and attacking targets without human control. Such systems can increase the risk of accidental conflict. Unlike earlier weapons, AI systems with dangerous capabilities could be easily proliferated through digital means. Also, AI trained to develop drugs could readily develop biochemical weapons. Chat GPT-4, for example, was able to conduct experiments and synthesize chemicals on its own in a real laboratory
(Global) conflict and wars are the third “C”
They are another source of global danger. Infrastructure wars are one example of this. As society becomes increasingly digitized and interconnected, the risk of cyberattacks that can cripple countries’ infrastructures or steal sensitive data also increases. Moreover, nuclear conflict remains a serious threat. Nuclear weapons could be used not only in warfare between nations but also in terrorist attacks – the scale of which could increase exponentially through the use of AI.
However, an analysis of possible armed conflicts can no longer be limited to geopolitics. Space is increasingly becoming an extension of terrestrial geography: geopolitics is being supplemented by astropolitics. By 2050, up to 250,000 satellites could be orbiting in space. With such numbers, the probability of collisions increases. If the quantity of debris keeps growing, Earth’s orbit will become impassable. Technologies that can decimate space debris also have military applications: they could eliminate active satellites.
The U.S. operates nearly two-thirds of all active satellites. China’s impressive space program is catching up and is under their military’s control. China’s rise could lure the U.S. into the so-called Thucydides Trap, named after the historian who described the Peloponnesian War: The threat of a new adversary to an established power has historically led to war in 12 out of 16 cases.
As a result, the three “Cs” pose a significant threat to humanity.
No other form of government can overcome the three “Cs” like democracy.
Democracies have a better track record than dictatorships with regard to all three issues. In contrast to democracy, autocracy proves less capable of (better) solving humanity’s challenges when it comes to combating the three “Cs.”
Let’s start with climate.
Consider the example of China, whose totalitarian structure drove environmental degradation to levels so extreme that Beijing’s air quality became unfit to breathe and the health of its top party leaders was at risk. Even critics attest on almost all environmental issues, “Functioning democracies do somewhat less badly than dictatorships, and by no means worse.” The climate, and thus humanity, will only be saved by democratic means.
While the calculated CO2 emissions per capita in 2021 (Qatar; Authoritarian regime: 36 tons, Germany; Full democracy: 8t) were almost twice as high in the USA (Incomplete democracy: 15t) as in China (Authoritarian regime: 8t), the People’s Republic emits in total twice as much CO2 as the USA.
China is the world’s largest emitter of greenhouse gases, accounting for nearly one-third of all global emissions. (Just over one-sixth of the world’s population lives in China, but emissions are twice as much, at around one-third). While coal consumption is declining around the world, including in the USA and Germany, it is increasing in
China – which continues to expand its coal-fired power by building new power plants. More than half of the coal mined globally is mined and consumed in China.
Even though China has recently invested a considerable amount into the expansion of renewable energies and is striving to become the world’s technology leader in this sector, there is little to suggest that it will reduce its high dependence on coal enoguh to become the global market leader. In 2020, for example, China saw 29 gigawatts (GW) of new coal-fired power plants come online, while the rest of the world saw capacity decline by 17.2 GW. For the period of the 14th Five-Year Plan (2021-2025), it is planned that even more coal-fired power plants with a total capacity of 200 GW will come online. (By comparison, Germany had coal-fired power plants with a total capacity of 42.5 GW in 2020). Climate protection is being implemented in China, specifically where its economic challenges are concerned, such as overcoming energy bottlenecks. Often, the focus is not on combating climate change, but on reducing the symptoms – such as meteorological early warning systems for unusual weather events, the breeding of climate-resistant grain varieties, or irrigation technologies that are intended to help compensate for predicted yield reductions as a result of climate change. In urban development, the aim is to promote energy-efficient building and the construction of so-called “sponge cities”, which are designed to better cope with unusually heavy rainfall events and flooding and to store water during a drought.
China’s international climate negotiations seem to be less concerned with the climate than with cleverly asserting its own interests. For example when China allowed itself to be exempted from reduction commitments as a “developing country” or when almost 60% of the emission certificates that industrialized countries could credit to themselves for investments in climate-friendly projects in developing countries were realized through projects in China – such as Chinese wind farms. At the 15th World Climate Summit in Copenhagen in 2009, planned new regulations failed due in part to China’s blockade. In 2015, the People’s Republic then announced that it did not want to reach the peak level of CO2 emissions until 2030 and CO2 neutrality until 2060.
For comparison: while the second exemplary autocracy, Russia (authoritarian regime), also speaks of 2060 with nebulous announcements, the democracies USA, Japan, and the EU want to achieve CO2 neutrality at least 10 years earlier. In the race against catastrophe, a decade can be decisive. In the democracy of Germany, the year 2045 should be as much as 15 years earlier. In this detailed comparison, it should be noted that further greenhouse gases, such as methane, which is even more harmful to the climate, are not taken into account by China in its climate plan. Thus, on closer examination (contrary to what is propagated), the authoritarian regime is anything but efficient. Ecology remains a secondary concern for autocratic states.
Here, too, autocracies put their own interests before those of their citizens.
Many autocrats mostly serve themselves (and try to manipulate electoral processes so that citizens cannot vote them out). To distract, autocrats appeal to racist, sexist, xenophobic, or homophobic prejudices. The climate change-related Covid-19 sheds light on this tendency. Many autocratic rulers downplayed the pandemic, rejected scientific evidence, spread misinformation, and failed to take basic measures to protect the health and lives of their populations. An important and progressive development that surely worries many autocrats was the formation of broad democratic party alliances to look past political differences, while the totalitarian nature of the country allowed China to initially deny the pandemic (which cost countless lives worldwide). Instead, China punished whistleblowers. Again, policy correction came far too late. Ergo: Climate change (and Covid-19) are better tackled by democracies.
And what is the assessment regarding wars?
Statistically speaking, democracies wage fewer wars than autocracies and usually win them (1816-1992: 81 % to 19 %). Democracies forge joint war and defense alliances and suffer significantly fewer casualties in wars than non-democracies. The proposition that democracies do not wage wars against democracies has proven extraordinarily robust according to statistical analyses – because those that do are inconsistent with basic value patterns of democracies, and there are far more institutional controls and procedural barriers here before a country enters a war. Moreover, democracies are generally more prone to negotiation and compromise. Never has a true democracy overrun another democracy with war!
Regimes, on the other hand, have a strong and stable tendency toward conflict.
A large number of armed conflicts have taken place not just between nations, but also between autocratic regimes and sometimes against the citizens of their country. A regime’s rule is also internally violent and constitutes another form of war. On the other hand, the so-called (vigilante-avoiding) “monopoly of force of the state,” may be exercised within democracies only for purposes desired or approved by the people; it is bound by law.
Democracies will also be better poised to better navigate the challenges posed by artificial intelligence.
Their foundation lies in the principles of power transition, contrasting sharply with autocracies where power preservation dominates. However, those who prioritize maintaining power may view AI primarily as a tool to enhance control, potentially rejecting necessary regulations and oversight for AI development. The quality of AI is crucial. Here, the inherent traits of democracy shine through. Openness, a fundamental characteristic of free societies, fosters the development of superior AI. Because AI is a learning system, it improves as its access to data expands.
This perspective is exemplified by China’s approach with “Ernie,” a significant competitor to ChatGPT, which is openly accessible. However, regimes like China who are focused on controlling social narratives may restrict openness in ways that limit access to diverse training data, thereby hindering AI development.
[1] Franco, E. G. (2020), The Global Risks Report, World Economic Forum, National University of Singapore, Oxford Martin School, University of Oxford, Wharton Risk Management and Decision Processes Center, University of Pennsylvania, https://www3.weforum.org/docs/WEF_Global_Risk_Report_2020.pdf
[2] Joho, K. & Happel, S. (2015, 15. September). Was hinter dem “Clean Diesel”-Betrug steckt. https://www.wiwo.de/technologie/mobilitaet/vw-abgasaffaere-was-hinter-dem-clean-diesel-betrug-steckt/12349856.html
[3] https://unidir.org/files/2021-07/Table-Top%20Exercises%20on%20the%20Human%20Element%20and%20%20Autonomous%20Weapons%20Systems%20Summary%20Report%20-%20UNIDIR%20-%20Final.pdf
[4] https://www.nature.com/articles/s41586-023-06792-0 https://www.scinexx.de/news/technik/erstes-ki-gesteuertes-chemielabor/ https://www.chemie.de/news/1179877/autonomes-labor-beschleunigt-chemische-entdeckungen.html www.safe.ai/ai-risk; https://arxiv.org/pdf/2306.12001.pdf
[5] Fernyhough, J. (2019, 18. November). China slowdown threatens Paris climate goals. https://www.afr.com/policy/energy-and-climate/china-slowdown-threatens-paris-climate-goals-20191118-p53bh5
[6] World Coal (2010, 14. Juni). Bouncing Back. https://www.worldcoal.com/coal/14062010/bouncing_back/ Bundeszentrale für politische Bildung (2022, 30 August). Chinas Klimapolitik. https://www.bpb.de/themen/asien/china/512519/chinas-klimapolitik/
[7] Human Rights Watch (2022, 13. Januar). Die Zukunft für Autokraten ist düsterer als sie scheint. edit.hrw.org/de/news/2022/01/13/die-zukunft-fuer-autokraten-ist-duesterer-als-sie-scheint
[8] Merkel, W. (2006). Demokratie durch Krieg. (WZB-Vorlesungen, 14). Berlin: Wissenschaftszentrum Berlin für Sozialforschung gGmbH. https://nbn-resolving.org/urn:nbn:de:0168-ssoar-110152 https://www.humanistische-union.de/publikationen/vorgaenge/193-vorgaenge/publikation/demokratie-durch-krieg/
[9] Merkel, W. (2006). Demokratie durch Krieg. (WZB-Vorlesungen, 14). Berlin: Wissenschaftszentrum Berlin für Sozialforschung gGmbH. https://nbn-resolving.org/urn:nbn:de:0168-ssoar-110152
[10] Merkel, W. (2006). Demokratie durch Krieg. (WZB-Vorlesungen, 14). Berlin: Wissenschaftszentrum Berlin für Sozialforschung gGmbH. https://nbn-resolving.org/urn:nbn:de:0168-ssoar-110152