Search the archive.

BEIJING — The most momentous challenge facing socio-economic systems today is the arrival of artificial intelligence. If AI remains under the control of market forces, it will inexorably result in a super-rich oligopoly of data billionaires who reap the wealth created by robots that displace human labor, leaving massive unemployment in their wake.

But China’s socialist market economy could provide a solution to this. If AI rationally allocates resources through big data analysis, and if robust feedback loops can supplant the imperfections of “the invisible hand” while fairly sharing the vast wealth it creates, a planned economy that actually works could at last be achievable.

Nick Chaillan, the U.S. Department of State’s first-ever software chief,  in September 2021. He claimed that the United States had “no competing fighting chance against China in the next 15 to 20 years.”

China is currently winning the race toward global dominance in artificial intelligence. But how did China get so far ahead?

The findings above reinforce the importance of situating contributions AI can bring to union growth and capacity in terms of the specific and changing institutional and identity contexts unions face, rather than positioning them in relation to abstract notions of ‘bureaucracy’ per se (Martínez Lucio, ). In this case study, the app offered the potential to transform the capacities of the alt-labour network in a dramatic way, promising the possibility of an autonomous, self-contained organising space for workers that was not reliant on the good graces of a major technology company. The transformational promise of the app was very different for the traditional union in Australia (despite the fact that both entities shared identities as ‘organising’, rather than ‘servicing’ organisations) because its affordances had to be reconciled with, and configured in relation to, existing capabilities and structures of a union that was over one hundred years old. This institutional context meant that the app was, from a relatively early stage, necessarily viewed as a tool that had to be understood as one that would only ever augmenting and enhancing existing systems, rather than replace them (Rego et al., : 326). The case study confirms, too, the tenacity of pre-existing ‘communities of practice’ (Martinez Lucio et al., ) in shaping how digital initiatives develop. Significantly, both the digital and non-digital ‘home’ organising environments in the traditional union (represented by home care workers and casino workers) proved highly durable in the face of early efforts to shift the locus of organising, wholesale, onto the app. Shorn of the possibility of the app becoming an autonomous self-contained organising space, UWU had to wrestle with how it might be best integrated into its existing capabilities, which included both legal ‘servicing’ and ‘organising’ dimensions, dilemmas that did not arise in the alt-union context where no pre-existing ‘servicing’ facility existed.

For the chatbot to advance the unions’ purpose and identity as an ‘organising’, rather than ‘servicing’ union (Hodder and Edwards, : 847), a series of conscious steps had to be taken to re-orient it from the epistemic footing set for it by IBM, as a customer service tool for streamlining human interaction, to one that laid the ground for increased substantive human interaction in the future. Remaking the chatbot as organising tool required a reconceptualisation of multiple aspects of its programming, from the way it was described and promoted, to the way it interacted with members during ‘chats’ and contextualised data, to the way that the ‘backend’ labour of data-provision was allocated and to the way that ‘success’ was calibrated. In place of metrics which placed value on the chatbot’s ability to ‘problem solve’, the app was instead evaluated for its value as an tool for initial engagement, issue detection and workplace mapping, all of which formed foundations for future, human-led, organising conversations. This re-orientation was undertaken consciously and carefully through a process of value-led reflection, observation and continual dialogue with members. It did not arise mechanically as a consequence of any blunt ‘disruption’ of existing union systems by digital technology or unmanaged ‘clash’ between logics of network and hierarchy. The process of designing and reframing the app was a resource-intensive, rather than ‘labour saving’ one (Rego et al., ). Once reconfigured as an organising tool, however, the chatbot contributed to strengthening union power and capacity in several dimensions of the Lévesque and Murray () framework outlined below.

 sought to answer these questions by examining how two labour organisations, the US-based Organization  (OUR) and the Australia-based  (UWU) used an AI-enabled chatbot in their organising. The app, WorkIt, had been initially developed by OUR to organise a workforce of more than one million workers without the institutional backing of a formal trade union or any rights to enter the workplace. It was subsequently developed by UWU in Australia in casinos, home care and hospitality. The chatbot provided automated responses to workers’ text-based questions about their rights via their handheld devices, using natural language recognition capabilities. Questions that could not be answered by the chatbot within an acceptable range of certainty were diverted to a human responder, whose answers were fed back into the chatbot’s ‘brain’ to increase its capacity to answer similar questions automatically in the future. Three key issues stood out as things to consider for trade unions thinking about using AI to build power:

Off-the-shelf versions of IBM Watson were designed for commercial purposes, on the assumption that organisations would use it to ‘streamline’ and reduce human-to-human interaction in service transactions as much as possible. Under the organising model, though, unions have a different kind of relationship with their members. They wish to engage and support them in exercising power collectively in their workplaces, which means using technology to start and enrich conversations with members, not end them as quickly as possible. To ensure that the app supported that philosophy, UWU fundamentally reconfigured the chatbot. In place of an initial ‘information delivery’ approach, it changed the chatbot’s programming. Information was still provided to members in a useful way, but it was framed to reflect the fact that the rules of work, and the way they are applied, are always bound up in questions of power in the workplace. This approach set the scene for future organising.

But as AI advances, the question of whether AI is avoidable grows within our daily conversations, and, more specifically, about how can we balance the use of AI and humans in various industries. Several factors suggest AI is indeed inevitable. One of the most significant is the growth in computational power over the past few decades. Powerful processors and algorithms have enabled computers to perform increasingly complex tasks, from facial recognition to natural language processing. Additionally, there is an ever-increasing demand for AI applications in the healthcare, finance, and transportation industries.

As these industries (and several more) become more reliant on automation and data analytics, the need for AI solutions will only grow. Furthermore, the potential benefits of AI are significant. For instance, , with predictive analytics and machine learning models helping doctors diagnose diseases and develop personalized treatment plans. AI can also aid in improving transportation by enabling autonomous vehicles to navigate roads safely and efficiently, reducing the number of accidents and traffic congestion.

In the two brief months since its launch, has already become the fastest growing app in history. Being used to deliver everything from and essays through to a variety of sophisticated search interfaces, it’s been hailed as an industry disruptor that will make endless professions obsolete and even threaten to . Pretty impressive stuff for what is, essentially, a chatbot. ChatGPT and other new generative AI tools clearly represent a noteworthy evolution in the sophistication of AI as a technology and the capabilities that “off-the-shelf” solutions can offer. However, as with any trend that comes with so much fanfare it’s important to sort the facts from the hype. Results so far are certainly far from perfect, given the number of , suggesting that AI likely isn’t ready for world domination just yet. Still, the global AI market is projected to reach and the transformative technology was on every top executives’ lips at this year’s for good reason. The launch of ChatGPT has been a seminal moment and served at the very least to capture our collective imagination at where this technology may lead next. We already know that AI has near endless use-cases which will continue to morph alongside the sophistication of the technology, but what about the big picture? Here are three predictions on what hype surrounding ChatGPT signals for the future. Next, the launch of ChatGPT has shown us how many questions still exist when it comes to the implementation of advanced AI technology. The future of AI will touch on everything from economic theory and business management to public policy and law, meaning future graduates need to be better equipped with specialized courses that equip them with comprehensive theory that can manage the future of this technology. In good news, the wheels for this change are already in motion: Stephen A. Schwarzman donated $5 million to to help shape the future of AI in a way that benefits the public interest and a staggering $188 million to to support research into the ethics of AI. Such sizable donations like these highlight the importance of AI-related courses across all fields of study to help us develop regulation laws, competitive advantage regulations and everything in between as this transformative technology continues to take hold.

With the Singularity drawing closer, so does the outcome for the future of humanity: we must create a Strong, Friendly, Artificial General Intelligence to ensure humanity survives the century without harm from global warfare, nanotechnology, UnFriendly AI, or other existential risks. I picked this topic because I can’t think of anything more important for humanity right now, and while I’m fairly confident I’m not smart enough to contribute much in solving the problems involved, I nevertheless feel I can get a feel for them and spread the word.

Imagine a power[6] that does all of the following: it can kill the largest of animals on Earth, it can create and control fire, it can shift water and rock around and create ovens to melt other kinds of rock into different shapes, it can make shelters for surviving cold climates, it can create high pyramids and high towers, it can pave roads, tame other larger beasts to use on the roads, it can treat many different diseases and ailments, it can record history and have a memory longer than any other life-form, it can cause a radioactive nuclear explosion and obliterate land, it can escape the Earth and reach the moon, it can create machines that operate on electrical impulses to perform simple binary logic that can do anything from addition to manipulating molecules. It is a single thing that can do all these tricks, and it is just behind your eyes. Our squishy brain is the singular reason for all of humanity’s accomplishments. We are Intelligent, as a species, and the dumbest of us are far superior to the smartest of other species. When people think of intelligence, they think of Einstein instead of simply a human being. Our history isn’t composed of Einsteins, yet we’ve come so far regardless: clearly being human is what intelligence means, rather than being on top of an IQ scale that only applies to humans, and only loosely. The universe has been around for billions of years, and the most complicated thing it has

Biological diversity is the “infrastructure” that supports life on our planet: the air we breathe, the food we eat and the water we drink are only possible as long as we have healthy biodiversity – rich species and ecosystem diversity. Yet, this essential human lifeline is severely threatened.

The latest data shows that we are on the brink of crossing ecological boundaries and reaching tipping points in climate and ecosystems that might lead to an acceleration of planetary destabilization. The World Economic Forum’s 2018 Global Risk Report lists ecological collapse and biodiversity loss among the top 10 risks in terms of impact.

Collectively, our estimates suggest that a large share of employment and work is at least partially exposed to automation by AI, raising the prospect of significant labor savings. To assess the robustness of our estimates, we compare our baseline US estimate to a wider range of scenarios, including those in which AI can perform more or less difficult tasks than we assume in our baseline, and in which we relax our assumption that AI cannot assist with jobs which are primarily outdoors or physical (i.e., a scenario in which AI is complementary with robotics and existing machinery). Our scenario analysis suggests that the ultimate share of work exposed to automation could range from 15-35% (Exhibit 7, left panel), a range which is consistent with—but on the conservative side of—existing estimates in the literature (Exhibit 7, right panel). Our relatively conservative baseline primarily reflects our narrower focus on the impact of generative AI, in contrast to other studies which sometimes consider a wider range of related technologies (including robotics) that increase the scope for automation. Although the impact of AI on the labor market is likely to be significant, most jobs and industries are only partially exposed to automation and are thus more likely to be complemented rather than substituted by AI. In Exhibit 8, we assume for illustration that jobs for which at least 50% of importance- and complexity-weighted tasks are exposed to automation are likely to be substituted by AI, while jobs with an exposure of 10-49% are more likely to be complemented, and jobs with a 0-9% exposure are unlikely to be impacted. In our baseline, these assumptions would be consistent with 7% of current US employment being substituted by AI, 63% being complemented, and 30% being unaffected, though the ultimate effects will depend on how labor demand and occupational workloads evolve in response to partial labor savings in the majority of occupations. The large share of employment exposed to automation from generative AI raises the potential for a boom in labor productivity that significantly increases global output. There are two main channels through which AI-driven automation could raise global GDP. First, most workers are employed in occupations that are partially-exposed to AI automation and, following AI adoption, will likely apply at least some of their freed-up capacity toward productive activities that increase output. Academic studies confirm that workers at early-adopting firms experience higher labor productivity growth following AI adoption, with estimates generally implying a 2-3pp/year boost (Exhibit 9). While differences in the capability of generative AI relative to earlier vintages make it hard to extrapolate these results forward, they clearly suggest that generative AI can drive an economically significant increase in productivity. Second, we anticipate that many workers that are displaced by AI automation will eventually become reemployed—and therefore boost total output—in new occupations that emerge either directly from AI adoption or in response to the higher level of aggregate and labor demand generated by the productivity boost from non-displaced workers. Both of these channels have plenty of historical precedent. For example, information technology innovations introduced new occupations like webpage designers, software developers, and digital marketing professionals, but also increased aggregate income and indirectly drove demand for service sector workers in industries like health care, education, and food services. To demonstrate how technological innovation that initially displaces workers drives employment growth over a long horizon, in Exhibit 10 we show results from a recent study by economist David Autor and coauthors.5 Using Census data, they find that 60% of workers today are employed in occupations that did not exist in 1940, implying that over 85% of employment growth over the last 80 years is explained by the technology-driven creation of new positions. Exhibit 11 leverages a different academic study by economists Daren Acemoglu and Pascual Restrepo that decomposes changes in labor demand into contributions from productivity growth and technology-driven worker displacement and reemployment (among other factors) to show how the drivers of labor demand change over time.6 Technological change displaced workers and created new employment opportunities at roughly the same rate for the first half of the post-war period, but has displaced workers at a faster pace than it has created new opportunities since the 1980s. These results suggest that the direct effects of generative AI on labor demand could be negative in the near-term if AI affects the labor market in a manner similar to earlier advances in information technology, although the effects on labor productivity growth would still be positive. The combination of significant labor cost savings, new job creation, and a productivity boost for non-displaced workers raises the possibility of a labor productivity boom like those that followed the emergence of earlier general-purpose technologies like the electric motor and personal computer. These past experiences offer two key lessons. First, the timing of a labor productivity boom is hard to predict, but in both cases started about 20 years after the technological breakthrough, at a point when roughly half of US businesses had adopted the technology (Exhibit 12, left panel). Second, in both of these instances, labor productivity growth rose by around 1.5pp/year in the 10 years after the productivity boom started, suggesting that the labor productivity gains can be quite substantial (Exhibit 12, right panel). To estimate the boost to labor productivity in the US from widespread adoption of generative AI, in Exhibit 13 we sum up the implied labor productivity effects from direct labor cost savings, a productivity boost for non-displaced workers, and a composition effect from the reemployment of displaced workers in new positions. Our baseline analysis incorporates our key findings from above, including that about 7% of workers are fully displaced but that most are able to find new employment in only slightly less productive positions, that partially exposed workers experience a boost in productivity consistent with existing estimates (Exhibit 9), and that effects are realized over a 10-year period that starts around the time when roughly half of businesses have adopted generative AI. Under these assumptions we estimate that widespread adoption of generative AI could raise overall labor productivity growth by around 1.5pp/year (vs. a recent 1.5% average growth pace), roughly the same-sized boost that followed the emergence of prior transformative technologies like the electric motor and personal computer. This estimated boost to labor productivity growth is both admittedly quite large and highly uncertain. Exhibit 13 therefore also considers other plausible scenarios and shows that the boost to US productivity growth could easily range from 0.3-3.0pp depending on the difficulty level of tasks generative AI can perform, how many jobs are ultimately automated, and the speed of adoption:

In many ways, it is unfortunate that labor market policy has dominated our thinking about the AI economy. The main economic impact of AI is not about jobs or, at least, is about much more than just jobs. The main economic impact of these technologies will be how good they are. If the recent advances continue, AI has the potential to improve the quality of our products and our standard of living. If AI helps us diagnose medical problems better, improves our highway safety, gives us back hours of our day that were spent driving in traffic, or even just improves the 3 quality of our selfies, these are direct consumer benefits. These raise our real incomes and the economic studies valuing the improvements from quality and from new products tend to show their value is often extremely high (see the discussions in the volume of Bresnahan and Gordon, 1997 or the discussions over valuing “free” goods like Goolsbee and Klenow, 2006 and Varian, 2013). That’s a different way of saying that if AI succeeds, it will raise our productivity and higher productivity makes us rich. It is not a negative. Indeed, if AI succeeded in the way some fear, it would mean the exact reversal of the main problem facing growth in the last decade or more that productivity growth has been too slow. Indeed, it would decisively refute one of the central tenets of secular stagnationist thinkers like Gordon (2017) who argue that low productivity growth is a semi-permanent condition for the advanced economies because of the scarcity of path breaking ideas. Would that AI could change that equation. This paper will consider a few disparate thoughts about policy in an AIintensive economy (interpreting AI broadly to include a cluster of information technology based productivity improvements beyond just conventional Artificial Intelligence or Machine Learning). It will consider the speed of adoption of the technology—the impact on the job market and the implications for inequality across people and across places, discuss the challenges of enacting a Universal Basic Income as a response to widespread AI adoption, discuss pricing, privacy and competition policy and conclude with the question of whether AI will improve policy making itself. Taking the issue of job displacement first, the basic conclusion of the economists is that for the last hundred years, there have been massive amounts of job displacement yet the structural unemployment rate has not seemed to rise, much less trend toward 100%. Over time, people adjust. They move. They get skills. The long-run impact of labor saving technologies have overwhelming been positive for market economies. If the fear is that AI will replace low-skill jobs, it is a fact that tens or even hundreds of millions of low skill jobs were displaced by technology in previous years in a process very similar to the one we describe today. If the fear is that AI is different this time around because it will begin to replace types of jobs that have never been automated before like higher-skill or white-collar jobs the historical data indicate that those groups have been able to adjust to shocks and move to new sectors and new geographic areas easier than lower skill workers have. A critical issue is, of course, how fast the adjustment takes place/the speed of adoption of AI technology. The economy has proven quite capable of inventing new things for people to do over the long-run. Obviously, if change happens all at once, the adjustment problem is worst. Spread out over time, however, the adjustment can be manageable. Take the much discussed case of autonomous cars. There were about 3.5 million truck, bus, and taxi drivers in 2015 and suppose that every one of them were lost due to advances in self-driving car technology. If this loss takes place over 15 years, this would average a little over 19,000 per month and compare that to the fact that in 2017 the JOLTS data show that the economy generated about 5.3 million jobs per month (with 5.1 million separations per month). The complete elimination of every job in the sector would increase the increase the separation rate by less than four tenths of a percent. It would force drivers into new sectors and be disruptive to their livelihoods. But as a macroeconomic phenomenon, the impact would be small. If that loss happened in two years, the impact would be quite significant. So it’s worth considering what influences the speed of adoption and, certainly, a key determinant will be how good the AI actually is compared to people. But, many analysts seem to view that as the only thing that will determine adoption rates. It’s worth considering at least two other factors: prices and adjustment costs. First, many of these AI innovations involve significant capital outlays up front and that alone may slow their adoption for some time. Ride-share drivers, for example, by some measures can barely cover the cost of operating their cars (including depreciation, fuel, maintenance and insurance) at the price of cars now. AI enabled autonomous vehicles are likely to cost substantially more per car than conventional cars when they become available to the public. Will companies be willing to incur large up front to bypass paying drivers? It really depends on prices that we do not yet know. Second, ‘better’ doesn’t always mean faster adoption. Economists have shown automated stock picking through index funds superior to active management for decades yet people still hold trillions in inefficient, high-fee funds. Millions of people have mortgages with higher than market interest rates that they don’t refinance, cell phone data plans that do not match their usage, and so on. There are tens of millions of people that do not use the Internet. Inertia is a powerful force slowing the adoption of technology products and is certainly worth remembering if we want to predict something like how fast people will give up common behaviors like driving for themselves. Third, in an important sense, we know that AI can only be as good as its training sample and there are some very different types of customers in the country that may make the AI quality improvements much more fitting for certain types of customers than others. Microsoft created an AI program to learn from Twitter and see if it could create content that people would think was written by a human. They started it in the U.S. and had to shut it down almost immediately because it became so abusive and offensive. It mirrored what it saw online. Running the same program in China, where Twitter is heavily censored, it has performed well and not turned abusive. The attributes of the product and the ‘quality’ of the product depend on how relevant the training sample is to that customer. This is likely to influence the adoption rate of the AI technologies in different places. Again, think of the autonomous cars. Will we gather loads of information about driving in urban areas and on highways or in Silicon valley from the early adopters, tailor the product to their needs but then find that it doesn’t work as well for dirt roads or rural places or places without Bay area weather? Heterogeneous demand is the hobgoblin of the AI mind. Groups that differ most from the training sample will likely be the slowest to adopt the technology in part because it will be the least helpful to them. That may lead to another manifestation of the digital divide. In this sense, the rise of AI technologies is likely to make the problem of income and of geographic inequality even worse. To the extent that new AI technologies are expensive and tailored toward the training sample of adopters, it will be like having lower inflation and greater consumer surplus going to those groups (for discussions about differences in prices and innovation across income groups or for online buyers versus offline buyers, see Jaravel, 2017 or Goolsbee and Klenow, 2018). Government policy will face the potential of divisions along red state/blue state or high-education/low-education locations or high-income/low-income neighborhoods even more than it does today.

The launch of chat-based artificial intelligence has sparked debate about how the technology may change——the jobs we do. Now there’s evidence that using such tools at scale in the workplace can benefit employees. A from the National Bureau of Economic Research (NBER) finds surprising boosts to productivity, especially for workers who are new to the job. NBER researchers studied the use of an AI chat tool that gave guidance to customer support agents at a business software company. During conversations with customers, the tool delivered chat suggestions to the agents, who were free to use or ignore them while troubleshooting. It’s no coincidence that the researchers chose an AI tool built for support agents: The customer service industry appears to have one of the highest rates of AI adoption. In of businesses that deploy artificial intelligence, more than a quarter reported that AI figures in their customer service operations, outranking all other use cases. Looking at the data from more than 5,000 support agents, the NBER research found that AI assistance markedly improved their day-to-day work: It boosted productivity, led to better customer interactions, and strengthened employee retention. Workers who used the tool resolved almost 14% more cases over the course of an hour and were less likely to leave their jobs. The AI also improved how customers treated agents. “Customers often vent their frustrations on anonymous service agents and, in our data, we see regular instances of swearing, verbal abuse, and ‘yelling,’” the researchers write. But with the use of the tool, customer interactions became more positive. “AI recommendations, which were explicitly designed to prioritize more empathetic responses, may improve agents’ social skills and have a positive emotional impact on customers,” the study notes. It adds that eliminating negative interactions—and the emotional labor of handling angry customers—can protect workers from burnout or leaving their jobs, too. In perhaps the most interesting outcome, AI assistance was especially useful for newer or less-skilled employees. When using the AI chatbot, agents with two months of experience performed as well as those with six months of experience. The tool helped flatten their learning curve, the study found. “We find suggestive evidence that AI assistance leads low-skill agents to communicate like high-skill agents,” the researchers state, adding that the AI tool helped newer agents gain experience more quickly too. The overall benefits were also stronger within that group: Agents with fewer than six months of experience became even more productive and less likely to leave their jobs after using the chatbot. (On the other hand, the findings show, the AI tool helped more experienced employees considerably less with their work.) The paper concludes that generative AI can make work easier for employees, lifting both their productivity and job satisfaction. But AI doesn’t do the work alone, the NBER researchers stress. “Importantly, the AI system we study is designed to augment, rather than replace, human agents,” they write. In an age marked by anxieties that AI may automate us , it shows how the technology can work alongside people .

Although scientists have calculated the significant positive welfare effects of Artificial Intelligence (AI), fear mongering continues to hinder AI development. If regulations in this sector stifle our active imagination, we risk wasting the true potential of AIs dynamic efficiencies. Not only would Schumpeter dislike us for spoiling creative destruction, but the AI thinkers of the future would also rightfully see our efforts as the ‘dark age’ of human advancement. This article provides a brief philosophical introduction to artificial intelligence; categorizes artificial intelligence to shed light on what we have and know now and what we might expect from the prospective developments; reflects thoughts of worldwide famous thinkers to broaden our horizons; provides information on the attempts to regulate artificial intelligence from a legal perspective; and discusses how the legal approach needs to be to ensure the balance between artificial intelligence development and human control over them, and to ensure friendly artificial intelligence.

Our technology, our machines, is part of our humanity. We created them to extend ourselves, and that is what is unique about human beings. – Ray Kurzweil1

Attempts to block or “relinquish” [3, 12] molecular manufacturing research will make the world a more, not less, dangerous place [13]. This paradoxical conclusion is founded on two premises. First, attempts to block the research will fail. Second, such attempts will preferentially block or slow the development of defensive measures by responsible groups. One of the clear conclusions reached by Freitas [4] was that effective countermeasures against self-replicating systems should be feasible, but will require significant effort to develop and deploy. (Nanotechnology critic Bill Joy, responding to this author, complained in late 2000 that any nanoshield defense to protect against global ecophagy “appears to be so outlandishly dangerous that I can’t imagine we would attempt to deploy it.” [12]) But blocking the development of defensive systems would simply insure that offensive systems, once deployed, would achieve their intended objective in the absence of effective countermeasures. James Hughes [13] concurs: “The only safe and feasible approach to the dangers of emerging technology is to build the social and scientific infrastructure to monitor, regulate and respond to their threats.”

We can reasonably conclude that blocking the development of defensive systems would be an extraordinarily bad idea. Actively encouraging rapid development of defensive systems by responsible groups while simultaneously slowing or hindering development and deployment by less responsible groups (“nations of concern”) would seem to be a more attractive strategy, and is supported by the Foresight Guidelines [10]. As even nanotechnology critic Bill Joy [14] finally admitted in late 2003: “These technologies won’t stop themselves, so we need to do whatever we can to give the good guys a head start.”

3. Protecting human dominance through regulation or setting tailored goals to maintain human existence

Having a timeless and robust definition of AI is of paramount importance when thinking of regulating AI. One cannot regulate a certain subject without establishing a robust definition of what it regulates. The ambiguity of the definition of AI is mainly due to the “I”, “intelligence” of the AI. Concepts like “intelligence”, “consciousness”, “free will” and “soul” accompanying it are yet to have deterministic definitions although the greatest minds of our planet have tackled them for thousands of years (Burkeman, 2015).21

The European Parliament has approved the bloc’s landmark rules for artificial intelligence, known as the EU AI Act, clearing a major hurdle for the first formal regulation of AI in the West to become law. AI has become a key battleground in the global tech industry as companies compete for a leading role in developing the technology — particularly generative AI, which can generate new content from user prompts. The rules are the first comprehensive regulations for AI. What generative AI is capable of, from producing music lyrics to generating code, has wowed academics, businesspeople and even school students. But it has also led to worries around job displacement, misinformation, and bias. During a critical Wednesday vote, the Parliament adopted the AI Act with 499 votes in favor, 28 against and 93 abstentions. The regulation is far from becoming law, but it is likely to be one of the first formal rules for the technology globally. European Parliament members agreed to bring generative AI tools like ChatGPT under greater restrictions. Generative AI developers will be required to submit their systems for review before releasing them commercially. The Parliament also decided to hold firm with a ban on real-time biometric identification systems, as well as controversial “social scoring” systems. Human rights campaigners had expressed concern over an attempt by the European People’s Party to water down the ban. Lawmakers nevertheless pressed ahead with it and agreed to prohibit biometric surveillance from all public settings. The laws have huge implications for developers of generative AI models, such as the -backed OpenAI’s ChatGPT and ’s Bard. Jens-Henrik Jeppesen, senior director of public policy at Workday, said the AI Act aims to “build safeguards on the development and use of these technologies to ensure we have an innovation-friendly environment for these technologies such that society can benefit from them.” “Those are the right goals in my view,” he told CNBC after the vote. The next stage is for negotiators at EU institutions, such as the EU executive body and 27 member states. Earlier in the day, Github CEO Thomas Dohmke called on European regulators to listen to the private sector, as it pushed ahead with rules for AI. “We encourage the European Union and the US government to move really fast and listen to those that built the technology, not only in the commercial business, but also in universities, in the open-source communities,” Dohmke told CNBC’s Arjun Kharpal. It comes as countries around the world are looking to bring in rules and standards for AI. On Monday, U.K. Prime Minister Rishi Sunak made a bold pitch to make the U.K. . The government is also gearing up to hold a global summit on AI safety later this year.

The impact of AI is now at the center. In the case of the current writers’ strike, workers fear that their work in writing scripts will be automated, without need for their input. In short, writers are concerned that AI will eventually be used to produce literary material for a film or show. The WGA seeks to limit the use of AI as a tool to augment writers’ work as opposed to automate it by requiring that literary material be produced by, and credited to, a person. The WGA also seeks to limit use of unionized writers’ prior work to train AI systems to produce similar material. The Alliance of Motion Picture and Television Producers (AMPTP), a trade association representing studios and networks, has so far refused to be limited in its use of AI. Instead, it has proposed annual meetings with the union to further discuss changing use of AI technology Although AI in its current form is far from developed or sophisticated enough to replace writers at this stage, the expiration of the WGA’s contract with the AMPTP is the impetus for the current debate. As a general matter, the scope and duty of an employer’s obligation to bargain with the union depends on whether the negotiations are taking place during the term of a contract or otherwise, and whether the matter is a “mandatory” or “permissive” subject of bargaining under applicable law. Under the National Labor Relations Act (NLRA), unions and employers are required to bargain in good faith about “wages, hours, and other conditions of employment,” otherwise referred to as “mandatory” subjects of bargaining. With respect to these matters, an employer must bargain with the union about the decision to take a particular action. This type of bargaining is commonly referred to as “decision bargaining.” By contrast, employers are generally not required to bargain prior to making core managerial decisions pertaining to the operation of the business. These types of matters are generally referred to as “permissive” subjects of bargaining. Nonetheless, employers are required to bargain about the effects of major operation changes that will affect the terms and conditions of employment. This type of bargaining is commonly referred to as “effects bargaining.” During the term of a collective bargaining agreement, an employer cannot unilaterally change the terms and conditions of employment or the terms of the collective bargaining agreement without first negotiating with the union, and neither party is required to negotiate with respect to matters covered by the agreement. If a matter is not covered by the agreement, employers must negotiate with the union over mandatory subjects of bargaining and the effects of any permissive subjects of bargaining. Notably, employers may reserve the right within a collective bargaining agreement to make certain decisions related to the operation of the business during the term of the collective bargaining agreement without first negotiating with the union. These clauses are typically referred to as “management rights” clauses. Once a contract expires, employers must honor the terms of the expired agreement while negotiating with the union over the terms of a new agreement until the parties reach impasse. An employer’s decision to introduce new technology in the workplace has typically been considered a permissive subject of bargaining. Therefore, absent a provision in the applicable collective bargaining agreement, employers are generally not required to bargain with unions over the decision to introduce new technology in the workplace. Nonetheless, because of the potential of AI to change the dynamics of work, its use in the workplace could require an employer to engage in effects bargaining, which could—depending on the circumstances— involve discussions on issues like job security, opportunities for retraining and job displacement caused by the implementation of new technology. Because use of AI is likely a permissive subject of bargaining, the WGA is currently looking to add language to the applicable collective bargaining agreement to limit the use of AI. Given the increasing prevalence of AI across industries, other unions may follow WGA’s lead. Future labor negotiations are likely to include demands similar to those being made by the WGA, and such demands are likely to become commonplace as unions seek to adopt provisions in collective bargaining agreements that limit the use of AI. These discussions are particularly likely to emerge in creative and professional industries, and it is anticipated that there will be an increased demand for agreements that preserve creative control and utilize AI as a tool, not a replacement. Notably, as with prior WGA strikes, other labor unions representing employees in the entertainment industry have made their support for the WGA known, including the Screen Actors Guild—American Federation of Television and Radio Artists (SAG-AFTRA), whose contract with AMPTP expires at the end of June. Interestingly, on June 3, the Directors Guild of America (DGA) announced that it reached a tentative deal for a new three-year collective bargaining agreement with AMPTP, which would provide that AI is not a person and cannot replace the work performed by the members. AI has already begun flexing its power as a transformative force for both employers and unions, and its integration into workplaces is expected to reshape and redefine labor dynamics. Demands for AI to be restrained are expected to increase in the future, particularly in the creative and professional spaces; the writers’ strike will likely prove to be only the first of many union challenges to the implementation of AI. Issues such as job security and creative control will likely emerge at the forefront of labor negotiations. As AI continues its rapid evolution with no signs of slowing, employers in industries across the board must be ready and willing to embrace new opportunities and mitigate potential challenges with the aim of creating a future where both humans and AI can coexist in the workplace. Against this backdrop, employers of both unionized and nonunionized workforces should also be mindful of the changing legal landscape with respect to use of AI.

“If it’s applied ethically and responsibly in collaboration with trade unions and the workforce, then AI actually has the potential to impact our members’ lives in a positive way,” says Budd. But regulations might fall far behind this objective. The Trades Union Congress (TUC) issued a  to this effect in April, arguing that ministers had failed to introduce the necessary regulations to safeguard workers’ rights amid the rapid rise of workplace AI. 

The TUC, which represents more than five million workers across England and Wales, remains dissatisfied with provisions in the government’s , the UK’s post-Brexit replacement for GDPR, as well as the  and the recent  from the Department for Science, Innovation and Technology, which critics argue favours Big Tech. Prospect’s deputy general secretary Andrew Pakes is just as excoriating. “The government’s AI strategy is silent on workers,” he says. “It talks about innovation but not industrial change or investment in it.” 

The acceleration in the 21st century of automation, artificial intelligence (AI) and emerging technology in the workplace has required U.S. labor unions to create new playbooks to defend their members' interests. Union concerns center around three central innovations: automation, AI performance tracking and workplace monitoring.

McKinsey & Co. projects that by 2030, roughly 40 million U.S. workers, many of them union members, will have been replaced by robotics and automation. Unions are challenged in blunting this impact because an employer often may introduce new technology into its operations without union consent or bargaining.

While artificial intelligence is rapidly improving and some economists  the technology will put millions of workers out of jobs, labor unions are fighting against it. In bargaining sessions, AI is increasingly becoming a central sticking point, with organizers making the case that companies are shortsighted to replace knowledge workers with technology that can’t match human creativity and is riddled with errors and bias.

Some unions have recently made strides, such as Hollywood directors who struck a  Saturday with motion picture studios and garnered promises that they “will not be replaced” by artificial intelligence. It was one of the first concessions organized labor has gotten regarding AI protections.

The Levy Economic Institute calculates up to 16 million could take part in such a program today (including the unemployed, those working part time seeking full time work and individuals currently inactive who might move into the labor market). Given the federal government would have to pay $15 an hour for full time jobs, plus benefits equal to 20 percent of wages, total labor costs per worker would be $37,440 per year. That’s before the cost of the materials for the programs and administration of the program itself. Even assuming some opt for part‐​time positions, and ignoring the non‐​labor program costs, we are talking about a gross cost of up to around 2.4 percent of GDP, significantly higher than the existing Medicaid program (2 percent of GDP). The net cost on these assumptions will be lower, of course. People who take jobs will require less in welfare payments and pay some back in taxes. Some might wisely consider it a risk for their employment fortunes to be tied to the whims of politicians and their willingness to fund this program, and so remain in the private sector. But even taking this into account, and assuming the policy generates the macroeconomic bounty that the Levy researchers expect, they still think the annual net cost will be between 0.8 and 2 percent of GDP, with the program employing up to 10 percent of the workforce. That would in itself be a huge new commitment to finance at a time when the long‐​term fiscal outlook is already dire, and the short‐​term deficit already expected to balloon to over 5 percent of GDP in the coming years. In reality, the fiscal costs are likely to be much, much higher, and the economic welfare losses even more significant, because in the labor market and broader economy, a public jobs guarantee program would significantly crowd out productive private sector activity. This type of policy will radically alter behavior of both workers and businesses, and so the supply and demand for labor. The , among those who worked in 2016, 70+ million Americans earned under $32,500 (the full‐​time job guarantee salary would be $31,200). Yes, not all of these would seek out positions on the jobs guarantee program. But a large proportion would, especially those employed in uncertain roles with low levels of job security. In fact, some even paid more than $31,200 might consider leaving their jobs to pursue guaranteed roles if they perceive better working conditions or an easier worklife (asked under what conditions someone would be fired from such a role, the Levy Institute paper suggests that you would be sacked for failing to go to work, but that your performance would not be judged by “private sector ‘efficiency criteria’”, for example.) It’s not inconceivable then that over 25 percent of the labor force could find itself part of the scheme. This crowd‐​out is likely to be particularly acute in low productivity regions, and (ironically) after economic downturns. A nationwide jobs guarantee program paying $15 an hour will be particularly attractive to workers in low wage regions, and by setting a de facto wage floor the program will prevent private investment in regions on the basis of cheap labor. Though no doubt there would be some demand spillovers from well‐​paid jobs, the net consequence is highly likely to be weaker private sector job creation in poor regions, which has been the experience of countries such as Britain with a nationwide minimum wages and public sector national pay bargaining. Proponents of the scheme see “higher labor standards” as a good thing, but absent productivity improvements, policies which raise labor costs significantly will reduce the quantity of workers demanded. There’s good reason to expect the policy will reduce the efficiency and productive potential of the economy too. Taxes will eventually need to be raised to cover the net cost of the program. In infrastructure and care giving provision, costs will rise – because nobody would now work in these directly substitutable sectors for less than the wage and conditions offered in the job guarantee program. This will waste resources, and there’s highly likely to be overinvestment in lots of relatively low value ventures and programs to ensure workers are employed, especially given the explicit aim is to provide employment rather than deliver projects at low cost. Throwing resources at regions with higher levels of unemployment and after recessions too will work directly against market signals and deter the mobility of labor (in geographic and industrial terms) and capital to its most productive uses given prevailing market conditions. This is important: yes, employment is highly likely to have some positive externalities; but the real driver of better living standards over time are productivity improvements, discovered by market‐​based activity. Proponents of this policy seem to put an enormous weight on the idea that time out of the labor market has huge scarring consequences which could be ameliorated by any type of temporary employment. But the literature on this shows that temporary jobs  to improve longer‐​term labor market outcomes. As if all these consequences were not bad enough, such a program will be ripe for corruption and political interference at the government, provider and individual level. Senator Sanders’ plan would be administered by the Department for Labor, with local and state governments submitting projects to regional offices for consideration. There’s a huge question mark on whether projects will be considered on economic grounds, when there might be an incentive for make‐​work schemes to aid particular politicians or indeed to put resources towards “public good” causes or NGOs more in line with the ethos of the governing party. For Democrats this might be for environmental issues. For Republicans it might be, say, for a wall on the southern border. NGOs and local public bodies themselves will have incentives to apply for federal funds for projects that would otherwise have occurred anyway, and to maximize the number of applications. Pork barrel projects would proliferate. What is more, at the individual level, the guarantee coupled with the purported unwillingness to judge worker performance on a commercial basis will incentivize low levels of work effort on the margin.