Search the archive.

On the first factor, measured productivity has increased rather slowly in recent years, even as the world seems to be captured by AI fever. 3 If AI‐related innovations enter the economy at the same slow pace as suggested by recent productivity statistics, then the transition will be slower than e.g. the wave of mechanization in the 1950 – 1970s, and the resulting disruptions may not be very significant. However, there are three possible alternatives: First, some suggest that productivity is significantly under‐measured, for example because quality improvements are not accurately captured. The best available estimates suggest that this problem is limited to a few tenth of a percentage point (see e.g. the discussion in Groshen et al., 2017). Furthermore, there are also unmeasured deteriorations in productivity, e.g. declines in service quality as customer service is increasingly automated. Secondly, the aggregate implications of progress in AI may follow a delayed pattern, similar to what happened after the introduction of computers in the 1980s. Robert Solow (1987) famously quipped that “you can see the computer age everywhere but in the productivity statistics.” It was not until the 1990s that a significant rise in aggregate productivity could be detected, after sustained investment in computers and a reorganization of business practices had taken place. Third, it is of course possible that a significant discontinuity in productivity growth occurs, as suggested e.g. by proponents of a technological singularity (see e.g. Kurzweil, 2005).

On the second factor, the disruptions generated by AI‐related innovations depend on whether they are labor‐augmenting or labor‐saving, using the terminology of Hicks (1932), i.e. whether at a given wage, the innovations lead to more or less demand for labor. Some suggest that artificial intelligence will mainly assist humans in being more productive, and refer to such new technologies as intelligence assisting innovation, IA, rather than AI. Although we agree that most AI‐related innovations are likely to be complementary to at least some jobs – e.g. the ones applying AI to solve problems – we believe that taking a broader perspective, progress in AI is more likely to substitute for human labor, or even to replace workers outright, as we will assume in some of our formal models below.

Two sorts of bias are worthy of highlighting in the context of agriculture.

First, initially at least, the main markets for AI tools for agriculture will be the United States, Europe, and (perhaps) China. There is a danger that farmers in other regions will inherit AI systems trained on data about crop yields or soil chemistry in the US, Europe, or China, rendering them highly problematic in the local context elsewhere (Keogh & Henry, ); efforts to correct this by retraining the systems using data about crops grown under local conditions will still embed a preference for some crops over others.

The rising use of AI in agriculture could result in the overuse of pesticides and fertilizers, contaminating the soil and water, and harming biodiversity. Implementing AI in agricultural practices to increase yields at the expense of maintaining ecosystem health could lead to monocultures and biodiversity loss. 

Using AI for environmental management also raises ethical questions. Decisions made by AI systems could be biased if they were presented with inaccurate or incomplete data. For instance, if an AI system received instructions to value economic growth over environmental protection, it might choose to put short-term financial gain ahead of environmental sustainability.

In conclusion, the water footprint[6] of AI models, including Chat GPT, is a significant environmental issue that requires attention and action from the AI industry. While the water[21] consumption of AI models is relatively small compared to other industries such as agriculture or energy production, it is still a concern due to the growth and increasing complexity of AI models.

Several solutions can be implemented to reduce The water [9]footprint of AI systems, including improving the energy efficiency of data centers, using more efficient algorithms, and adopting water recycling technologies. These solutions can help reduce the amount[15] of water used by Chat GPT and other AI models, thus reducing their environmental impact.

Volume 156, November 2021, Science Direct, //DH

Rockström et al. (2009) warned that chemical pollution is one of the planetary boundaries that ought not to be crossed to safeguard humanity. Altogether more than nine million humans are dying prematurely each year – one in six deaths – due to contamination of their air, water, food, homes, workplaces, or consumer goods (Landrigan et al. 2018). To place this in perspective, the chemical-related annual death toll is significantly greater than that of World War II and today constitutes the greatest preventable form of mortality. Furthermore, it inflicts catastrophic losses on wildlife, notably insects and animals that depend on them, ecosystems and their services, such as pollination or clean water, on which humans depend for our own existence. This underlines the role of chemical pollution in potential planet-wide ecological breakdown (Dave 2013). There is increasing evidence in recent decades of cognitive, reproductive and developmental disorders and premature deaths caused by chemical contamination of the human living environment (Diamanti-Kandarakis et al. 2009).

Global water scarcity can have a severe impact on the economies of many, if not all, nations. One area that will be impacted is bilateral international trade. For those nations experiencing water scarcity they will tend to produce less agricultural goods and fewer manufactured products and will have to import greater amounts of each but at significantly higher prices (Hertel et al., 2013). In this case, these nations will be at a severe disadvantage since they will have to borrow more funds at higher interest rates thus causing financial deficits that will not improve over time. This will also cause a dangerous imbalance in the global trade market causing certain nations to go deeper into debt while others become wealthier over time. Financial projections state that huge food and agricultural trade deficits due to water stress could occur in South Asia at $1.35 billion while the Middle East and North Africa (MENA) region will have a deficit of $0.6 billion and nations such as China will see a gap of $1.08 billion and India of $0.44 billion (Hertel et al., 2013). In this case, trade occurs in the water used in the process of the good or product and is known as virtual water (Klepper, 2007). Currently, there are approximately 1,000 km3 of virtual water internationally traded every year (Klepper, 2007). Another area of economic impact is to businesses worldwide. Industries and businesses use large amounts of water in order to manufacture goods and products and in certain cases to provide services. For example, Chile’s copper production uses substantial amounts of water and its key mining area is in the northern region next to the Atacama Desert which is regarded as one of the most arid areas on the planet. This has caused water to not only become a more expensive commodity but also one that local farmers are battling for water rights (Cartwright, 2009). Many businesses are now taking into consideration the availability of water as a vital natural resource when making decisions on locations for investments or manufacturing. Water for numerous companies was taken for granted for many decades since it was abundant, cheap, and generally clean. But now due to threats of increasing water pollution and water scarcity on a global scale, many firms are paying more attention to water access than ever before. This also has a ripple effect on municipalities since the loss of firms due to water scarcity means declining tax revenues, job loss, and shrinking economies.

Water withdrawal refers to freshwater taken from the ground or surface water sources, either temporarily or permanently, and then used for agricultural, industrial or municipal uses. On the other hand, water consumption is defined as “water withdrawal minus water discharge”, and means the amount of water “evaporated, transpired, incorporated into products or crops, or otherwise removed from the immediate water environment”. While the evaporated water still stays within our planet just like any other matter, it may go somewhere else and further contribute to the already uneven distribution of global water resources. By default, water footprint refers to water consumption. But water withdrawal is also a crucial metric and indicates the level of competition and dependence on water resources among different sectors. Indeed, electricity generation is among the top sectors for water withdrawal in many countries. The global AI demand may even require 4.2 – 6.6 billion cubic meters of water withdrawal in 2027, which is more than the total annual water withdrawal of 4 – 6 Denmark or half of the United Kingdom. If the U.S. hosts half of the global AI workloads, the operation of AI may take up about 0.5 – 0.7% of its total annual water withdrawal. Simultaneously, the total scope-1 and scope-2 water consumption of global AI could exceed 0.38 – 0.60 billion cubic meters, i.e., roughly evaporating the annual water withdrawal of half of Denmark or 2.5 – 3.5 Liberia.

In section 3, we decompose the mechanisms through which innovation leads to inequality into two channels. First, inequality rises because innovators earn a surplus. Unless markets for innovation are fully contestable, the surplus earned by innovators is generally in excess of the costs of innovation and includes what we call innovator rents. We discuss policies that affect the sharing of such rents, such as antitrust policies and changes in intellectual property rights. The second channel is that innovations affect market prices; they change the demand for factors such as different types of labor and capital, which affects their prices and generates redistributions. For example, AI may reduce a wide range of human wages and generate a redistribution to entrepreneurs. From the perspective of our first- best benchmark with complete insurance markets, these factor price changes represent pecuniary externalities. We discuss policies to counter the effects of the resulting factor price changes.

4. Labor Harms. At least since the Industrial Revolution, people have been anxious about technology’s impact on workers—that they would be displaced by machines or, perhaps worse, turned into them.246 While most scholars believe that productivity gains from innovation outweigh the costs to workers in the long run, some recent research has been less sanguine.247 Artificial intelligence innovations may be displacing workers and suppressing wages at higher rates than in previous eras.248 In addition, new developments in worker surveillance are threatening workers’ health, privacy, and well-being.249

It is natural for innovation to make some jobs obsolete. The widespread adoption of automobiles had a devastating effect on buggy whip manufacturers.250 And while fewer people are needed to build a car than in the 1960s, the social benefits from productivity gains have historically been valuable.251 As robots increasingly replace human workers, however, displacement effects may exceed productivity gains. Recent research shows that increased use of robots, especially in manufacturing fields, significantly depressed both wages and employment.252 Economists now worry that the potential productivity gains from automation may not be strong enough to outweigh losses to labor. Darren Acemoglu and Pascaul Restrepo have recently warned of the possibility of excessive automation which “could reduce GDP and welfare.”253

Another dark side of innovation relates to end-of-product-life considerations, that are predictable from the start, but often ignored until the end. Examples include the disposal  of nuclear waste, mining clean-up operations, and heavy metal leakage from the recycling  of electronic waste. With new technologies, and especially technologies with long working lives, these end of life costs may be uncertain. In some instances, a new technology  can reduce these costs, but in many others rising social expectations and more stringent  environmental and health regulations can substantially increase costs over time.  Recycling can reduce the environmental impact of waste. However, if new materials  such as plastics are relatively cheap, recycling can be prohibitively expensive compared to  landfill or burning. In one example, from the UK, Westminster council sent 82% of all  household waste – including that put in recycling bins – for incineration in 2017/18  (Franklin-Wallis 2019). Disposal problems are exacerbated in the case of planned  obsolescence, where products have deliberately shortened lifetimes, or are designed  such that they are impossible or difficult to repair.1 For example, Apple has been  repeatedly accused of deliberately slowing its iPhones in order to increase the sales of  new products, and restricts the repair of Apple products even in the case of fairly simple  repairs (Svensson et al. 2018).  

Despite the usual focus on the benefits of renewable energy, Hansen and Nygaard (this  issue) provide evidence on the dangers. New technologies for solar panels cause problems  of heavy metal disposal, in particular for lead, which is a potent neurotoxin. Many Sub-  Saharan African nations lack capabilities to dispose of waste, and there are uncertainties  about the longer-term environmental performance of solar panels. The recycling of  hazardous materials is often done by the poor who lack equipment, protection, and  political representation. Innovations that address one kind of pollution can create new  problems, that fall on different people, and these costs should not be ignored, even if the  overall improvement is substantial.  

2. Environmental Harms. It is abundantly clear that innovations can harm the environment. Scientists attribute a substantial portion of global climate change to human activities, many of which are the result of previous innovations.213 Obviously, coal-based electricity plants and gasoline-powered internal combustion engines are responsible for an enormous share of environmental pollution.214 This is not to say that electricity and automobiles have been a net negative for society. We don’t think so. But there are many innovations that have generated substantially fewer benefits while producing meaningful environmental degradation. Sometimes, innovations can harm the environment directly. For example, hydraulic fracturing, also known as fracking, is believed by many scientists to cause increases in air toxins and substantial harm to groundwater.215 In addition, mining scarce minerals that are required for contemporary computing devices may be destroying environments in already ecologically stressed parts of the globe.216 And, of course, many innovations generate enormous amounts of air and water pollution.

Innovations don’t only harm the environment directly, however. In a range of industries, innovations contribute to environmental harms through planned obsolescence, the phenomenon in which people are encouraged to discard perfectly useful items once a new version is released.217 For example, over 100 million cell phones and over 300 million personal computers are discarded each year.218 Millions of new televisions are purchased, but very few are refurbished.219 Many of these products are still fully functional, but consumers have been induced into desiring new versions of them.220 Often, the new version may not represent any real upgrade in the product’s quality.221 In other cases, manufacturers intentionally limit a product’s lifetime in order to encourage new purchases.222 For example, light bulb filaments have been designed to burn out more quickly than necessary, and smartphone companies have designed software updates that degrade the phone’s battery.223 Finally, manufacturers can shorten product lifetimes by limiting consumers’ ability to repair their devices.224

Cluster A is in a central position in the bibliographic coupling network in Figure 3. Research included in this cluster studies the adverse consequences of the adoption of Information and Communication Technologies (ICT) in organisations – with a particular focus on pervasive ICT technologies (i.e. technologies that can be used almost anytime and anywhere). The adverse consequences considered relate, mainly, to psychological effects on employees after technology adoption, related to job satisfaction and turnover intention (Ragu-Nathan et al. 2008), work-life conflicts (Golden 2006) and burn-out (Maslach, Schaufeli, and Leiter 2001). For instance, Turel, Serenko, and Giles (2011) provide an empirical analysis of North American firms and show how addiction to workplace technologies (e.g. mobile technologies; email) can decrease productivity and organisational commitment and increase turnover intention and work–family conflicts. The addiction symptoms identified refer to employees’ inability to reduce their use of workplace technologies due to psychological dependency. In this context, Turel et al. show that these employee additions threaten their social lives (family and friends) and cause self-neglect. Golden, Veiga, and Dino (2008) find similar results for the impacts of remote working technologies, among large high-tech companies that employ remote workers permanently. The authors identify a measure of professional isolation to investigate its connection to job performance and turnover intentions. The more recent studies in this cluster focus on the impacts of workrelated social media use on employees. For instance, van Zoonen, Verhoeven, and Vliegenthart (2016) argue that work-related social media use can trigger role pressures that have negative psychological effects leading to emotional exhaustion among employees.  

[Marko; August 11; Lecturer at the university of helsinki’s department of social research and author; The Digital Age and it’s Discontents, “The Blind Spots of Digital Innovation Fetishism,” Ch. 2, P. 40-41] TC

Those who want to defend high-level military and intelligence spending often claim that military R&D has positive economic effects, since it creates spin-off technologies for the civilian sector, as in the case of advanced defence research projects that gave birth to the Internet. There is no need to disprove this argument completely, but it is very problematic from the perspective of democracy and citizen well-being. Indeed, one of the key reasons why the US elites have favoured military Keynesianism and its continuation in various forms is that massive state spending on the military is a better alternative for them, for reasons of power and privilege, than spending on social welfare. This is because social spending may

. Accessed 6/25/24.] LT

First, in order for financial innovation to have a positive effect on financial intermediation, it must enable an economically productive usage of savings that would not otherwise occur. If a family is willing to pay $300,000 for a new house that will cost $250,000 to build (including land), and they can pay off a loan comfortably over 30 years, then buying a house is an economically productive usage of savers’ resources that would not occur if mortgages did not exist. But the financial innovation (a mortgage) does not make the world better in and of itself; it depends on someone else’s having found a useful way to employ financial resources. Second, financial intermediation creates value when it moves funds to places where they are more valuable—making credit more available. But it is possible for the economy to be in a state where anyone who can employ credit effectively already has access to credit—or where people arguably have too much access to credit. With the benefit of hindsight, it is easy to see that the U.S. housing sector passed this point earlier this decade. With negative-amortization mortgages (where the monthly payment was less than the interest, causing the principal to go up) and stated-income loans (where the loan originator did not verify the borrower’s income), virtually anyone could buy a new house—leading developers to build tens of thousands of now empty houses, their current value far less than their cost of construction. In short, excess financial intermediation can destroys value when it causes people to make investments with negative returns; in that case, the banks would have been better off taking their deposits and investing them in Treasury bills. So in evaluating financial innovation, we need to think about whether it promotes beneficial financial intermediation or excessive and destructive financial intermediation. We cannot say that innovation is “good” simply because there is a market for it. The fact that there was a market for new houses does not change the fact that building those houses has turned out to be a destructive use of capital.

. Accessed 6/25/24.] LT

At the height of its power, in the run-up to the crisis of 2008, much of the influence of the financial sector was based on the fact that bankers were thought to be smart. “Financial innovation” was believed by many policy makers to be just as profound and just as conducive to productivity growth as the steam engine or the telephone had been in the nineteenth century. Under the approving eye of regulators, big banks were allowed to become larger—and to take risks that were bigger relative to their balance sheets. As a result of the financial-innovation-based boom and the crisis, the remaining big financial firms became bigger and remain very powerful politically. At the end of the fourth quarter of 2010, the six largest U.S. bank holding companies had assets valued at just over $9 trillion, which was 63% of gross domestic product (GDP). This is up from around 55% of GDP before the crisis (e.g., 2006) and no more than 17% of GDP in 1995. In the mid-1990s today’s six largest banks had assets around 20% of all banks insured by the Federal Deposit Insurance Corporation (FDIC). Today the size of these same banks, measured in the same way, is up to about 60%. At least roughly, this gives an indication of how the largest banks have risen in importance to the overall financial system—partly because of how financial innovation has played out. With assets ranging from around $800 billion to nearly $2.5 trillion, these bank holding companies are perceived by the market as “too big to fail,” meaning that they are implicitly backed by the full faith and credit of the U.S. government. They can borrow more cheaply than their competitors and hence become larger. Despite reform efforts since the crisis, including the Dodd-Frank legislation, big banks today still pose major structural system risks. Neil Barofsky, the special inspector general for the Troubled Assets Relief Program (TARP) summarized the situation well in his January 2011 report, emphasizing “perhaps TARP’s most significant legacy, the moral hazard and potentially disastrous consequences associated with the continued existence of financial institutions that are ‘too big to fail’” (SIGTARP 2011, 5). In public statements, top executives in these very large banks discuss their plans for further global expansion—presumably increasing their assets further while continuing to be highly leveraged. The next set of proposed innovations appears to be intensely cross-border in nature. There may be important and significant risks associated with these expansions. A very large cross-border bank liquidation cannot be handled through ordinary bankruptcy or through the currently operating resolution mechanisms (see the discussion below). Nevertheless, this global banking strategy is endorsed or at least supported in public by senior leadership at the Treasury Department (Scheiber 2011).

. Accessed 6/25/24.] LT

There has been a great deal of financial innovation in recent decades but its social value is unclear. In the run-up to 2008, banks took large amounts of risk relative to the size of the economy. This approach was made possible by and sometimes justified in terms of “innovation.” But it also created a great deal of downside risk for the economy—including widespread job losses and a big increase in the fiscal deficit.

[Marko; August 11; Lecturer at the university of helsinki’s department of social research and author; The Digital Age and it’s Discontents, “The Blind Spots of Digital Innovation Fetishism,” Ch. 2, P. 41-42] TC

Besides assisting in the creation of new financial innovations, digital technologies are at the heart of current global financial market infrastructure. Financial trading is digitalized, and an interesting innovation in this regard is high-frequency trading (HFT). This differs from earlier electronic forms of financial trading in that, instead of direct human involvement, trading is conducted fully automatically via computer algorithms and programs. HFT represents ‘the culmination of decades of technological innovation and regulatory developments encouraging financial automation’ (Lange, Lenglet & Seyfert 2016: 154). The appeal of HFT is based on its ability to execute financial orders at very high speed and to detect small trading opportunities in the markets. HFT systems do this by comparing available prices and price fluctuations between exchange venues and by taking advantage of temporary price inconsistencies to make profits (Bajpai 2014). HFT involves small individual trades in high volume, which has significantly increased the overall volume of financial trading in the 2000s (Chaparro 2017). The quantities are awe-inspiring: Typically, messages concerning bids to buy or sell shares are sent by computer algorithms at a rate of millions per second between the main US share-trading data centres. According to current estimates, about half of US share trading is done via HFT (MacKenzie 2018: 1636–1637).

[Marko; August 11; Lecturer at the university of helsinki’s department of social research and author; The Digital Age and it’s Discontents, “The Blind Spots of Digital Innovation Fetishism,” Ch. 2, P. 40-41] TC

ICT = Information (and/or) communication technology

Cluster B displays significant growth in the number of publications per year, compared to the other clusters. Most of the articles of this cluster have a meso-level unit of analysis and examine socio-technical systems (e.g. electricity, agro-food, transportation, etc.) These works address the unintended and harmful effects of innovation and investigate the complex interactions between the biophysical, social, economic and institutional factors resulting from the introduction of an innovation in the current socio-technical system. These studies tend to emphasise the directionality of transitions, arguing that innovations can have a tendency to develop in unsustainable ways. For instance, in an ethnographic study of Danish households, Røpke, Christensen, and Jensen (2010) explore unsustainable patterns of consumption of ICT. The study shows that the energy that is consumed by these technologies (including the most energy efficient ones) keeps increasing with their incorporation in everyday social practices. As a result, the energy reductions derived from more efficient technologies are outweighed by the energy consumption of ICT in households. In a study of the agricultural system, Wigboldus et al. (2016) investigate the scaling of agricultural-improving technologies such as disease-resistant seeds, permaculture cultivation practices and automated milking systems. They show how the scaling up of these technologies may pose a series of undesirable effects such as unanticipated environmental degradation effects, poor farm worker labour conditions and loss of control by farming communities over seeds. For example, pesticides give positive economic returns in the short run, but generate negative externalities in the form of environmental degradation and adverse impacts on human health. The adoption of these detrimental innovations is ascribed (mostly) to the fact that the agricultural system became locked into a path-dependent production model which has developed along a direction that it is difficult to change despite its being environmentally and socially unsustainable. 

Of particular interest is that the research in this cluster, stresses that, in some cases, even innovations designed to contribute positively to the environment – see, e.g. improvements to materials and improved energy efficiency in buildings – may cause significant unintended negative impacts on society, including greater consumption and increased waste and pollution, giving rise to the so-called ‘rebound (or take-back) effect’ (Frondel 2004; Herring 2004). Herring and Roy (2007) study is a fitting example of this. They examine the rebound effect in an empirical study focused on the factors that influence consumers’ adoption of low and zero carbon technologies. They show that adopting energy-efficient technologies can result in reduced energy consumption, in the short term, of between 10% and 20%; however, this positive effect is soon outweighed by rising levels of consumption. Rebound effects are observed, also, in the transportation system, as noted by Wang, Zhou, and Zhou (2012), who investigate the direct rebound effect for passenger transport in China in the period 2002–2007 and show that energy-saving transport systems reduce transport costs, but that the energy reductions are offset by increased demand for public transport due to its lower costs. 

It is true that the political agenda has been oriented towards stimulating economic competition, wealth creation and productivity through technological progress, which, in turn, have improved some of the economic and social indicators in most market economies. However, at the same time, innovation-induced industrial activities have had important negative consequences, in terms of both environmental disasters and deterioration of human rights (Giuliani 2018). Indeed, modern economic growth is based on intensive use of non-renewable energy sources, industrial chemicals and the production of massive non-recyclable waste. For instance, Luzzati, Parenti, and Rughi (2018) show that increased incidence of cancers is becoming a public health issue in most of advanced economies, allegedly due to their population’s exposure to the toxic emissions of industrial activities. Use of chemical inputs, such as pesticides, are another example of failed ‘economic progress’ as a result of technological change; it strikes that the traces of 12 of the most toxic chemicals ever invented, which were banned after the 2001 Stockholm Convention for their noxious impacts on human health and the environment, can still be found in the most remote areas of the world, such as deserts, open ocean and arctic regions, where there is no human activity or chemicals applications (Lohmann et al. 2007). Innovations have also had perilous consequences in finance, for their calamitous contributions to the 2008 financial crisis (Dosi 2011, 2013): financial innovations such as sub-prime mortgages, collateralised debt obligations (CDO) and credit default swaps (CDS) all played a central role in creating the crisis, giving rise to a process of ‘destructive creation’ (Soete 2013). In another context, scholars have noted that, at times, technological change has also had detrimental effects on employment which have not been entirely compensated by the wave of opportunities coming from the change. For instance, Korinek and Stiglitz (2017) recently showed that the diffusion of artificial intelligence and other forms of worker-replacing technological progress, could have a negative impact on both income distribution and employment.