{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"# I.\t Introduction\n\nGenerative Artificial Intelligence (Generative AI) is a deep learning technology used to generate novel content, such as text, computer code, audio, video, and images (McKinsey & Co. 2023a). Examples include Large Language Models such as Google Bard and Open AI's ChatGPT-4. These models are currently receiving much attention, due to their advanced abilities to produce language and computer code--for example, ChatGPT-4 recently passed a bar exam (Byrnes, 2023). \n

\nAlmost all experts agree that generative AI technologies will profoundly affect our lives in numerous ways,\nthough how exactly is a matter of debate, as the technology's capabilities and adoption remain in an early stage. But many thought leaders (including Elon Musk, a founder of OpenAI) believe that it could lead us down the road to dystopia or even self-destruction (Begley, 2023). One commonly cited example, and this essay’s focus, is this transformative technology’s threat to jobs. Specifically, many believe that, at a massive scale, Generative AI will eliminate jobs entirely and/or render today's job tasks obsolete.\n

\nConcern about this threat is rooted in a pervasive underlying assumption--so pervasive that it need not be stated: Full-time wage paying jobs are desirable, and the more the better. For instance, every serious presidential candidate in recent history has promised “jobs, jobs, and more jobs”, and their opponents never respond by arguing that this is not a good idea—rather, they argue that their own policies meet this goal in a superior manner. \n

\nIn this essay, I offer an alternative perspective. I propose that authoritarian structures, such as those that prevail in a traditional workplace, whereby companies dictate to employees what to do and when, where, and how to do it, limit human potential, freedom, and general well-being. This point of view, once limited to libertarian thinkers considered unrealistic or radical, is gaining greater acceptance—as a society, we are coming to expect more individual freedom both inside and outside the workplace. Generative AI, as a technology that will provide companies and employees much greater flexibility regarding how and even if jobs are performed, is poised to greatly accelerate this cultural attitude shift.\n

\nBelow, I expand upon these points and review the current thinking around how generative AI will impact workplaces in the future. My review suggests that AI will eliminate many jobs and alter others beyond recognition. But rather than threatening dystopia or Armageddon, I conclude that this transformation, compared to the current jobs-based economy, could result in a net benefit to humanity, by providing greater opportunities for creative work and human freedom.","metadata":{}},{"cell_type":"markdown","source":"# II.\t The Full Employment Ethic and its Discontents\n\nAs noted in the introduction, there is no serious disagreement among contemporary leaders that, for many reasons--economic as well as social--\"jobs for all\" is a desirable policy goal. As one economist puts it:\n

\n

But beyond money, a job is also crucial for establishing a person’s sense of security and self-worth, health and safety, ability to raise a family, and chances to participate in the life of the community (Polin, 2011)
\n

\nThis is a good summary of a narrative that, because it is so widely accepted, no one needs to articulate explicitly. However, throughout history and across cultures, competing viewpoints held greater acceptance. For instance, the ancient Greek Aristotle believed that paid work “degraded human character” by diverting people away from leisure; the ancient Roman Cicero wrote that \"the very wage [hired workmen] receive is a pledge of their slavery\"; and in some Caribbean and African cultures, enterprises have been accused of paying witches to turn their kin into zombies, so they would be willing to work for wages. (Falk, 2021). \n

\nIn modern Western philosophy, libertarians such as Wilhelm von Humboldt, Jean-Jacques Rousseau, and more recently Noam Chomsky have critiqued and opposed “wage slavery”. To briefly summarize, they believed that “the essence of human nature is human freedom, creativity, and connectivity – that “man is fundamentally a creative, searching, and self-perfecting being” (Chomsky, 1987, p. 149) and that “human beings naturally strive for greater personal freedom and resist constraints on their autonomy” (Ibid., p. 143) Paid labor, because it is coercive and prescriptive in its very nature, and because it is based on a “concept of competitive man who seeks only to maximize wealth and power” (Ibid., p. 153) contradicts these essential qualities. \n

\n\nMany management theories, under which the value of paid work is a given and probably would otherwise frame the above ideas as misguided, have incorporated this emphasis on human freedom when it suits the purpose of increasing employee productivity. For instance, multiple theories identify and encourage practices designed to increase employees’ perceived levels of individual control over their work (Ryan and Deci, 2000), based on research demonstrating such perceptions are associated with higher job satisfaction and productivity (Spector, 1986). And Agile methodology, today's prevailing management approach to software development, recommends providing knowledge workers with “more freedom and autonomy to do their job” in order to “improve [their] quality, productivity, and happiness.” \n

\n\nThese workplace theories and practices basically hack the human need for freedom-—they offer more autonomy selectively and attempt to lead employees to perceive that they have control over their activities, as a means to benefit employers. But now we are in the midst of a cultural paradigm shift, whereby worker and individual freedom is valued to a greater degree for its own sake, as an “end” in itself. The COVID-19 pandemic provided momentum for this values shift—-for instance, it led to greater demand for remote work; the “great resignation” whereby people simply refused to do unfulfilling work merely to meet their minimal physiological needs; and expectations of a work-life balance weighted more towards life outside work. (Moss, 2022). And now, almost all of sudden, just as people expect more time for themselves away from work, and less time while at work engaging in drudgery, comes generative AI, a technology highly suited to delivering on these expectations.\n","metadata":{}},{"cell_type":"markdown","source":"# III.\t The Impact of Generative AI on Employment\n\nTo help understand how generative AI might change jobs in a way that engenders a shift in values, we need to deep-dive into ways the technology might change the employment landscape. I will survey the relevant findings and opinions of published studies and writers, and, as someone who has spent lots of time reading these publications and thinking about the issue, offer my own thoughts.\n\n## A.\tExperts Agree: Generative AI will Perform Many Job Tasks.\n\nExperts as well as the results of studies support the prediction that generative AI will be able to replace human workers on a large variety of job tasks. One study (Goldman Sachs, 2023) finds that 50% of total job task hours could be automated by generative AI and other technologies; another puts this figure at 60-70% (McKinsey & Co. 2023b). \n

\nAs for the types of tasks that generative AI could perform, in the words of one author, \"just about every job that requires creativity or working with information is likely to be affected in some way\" (Marr, 2023). The McKinsey report cited above provides examples such as \"generative AI’s ability to support interactions with customers, generate creative content for marketing and sales, and draft computer code based on natural-language prompts\" and identifies four business functions--customer operations, marketing and sales, software engineering, and research and development--that could account for 75% of the tasks replaced.\n

\nAs illustrated by the below figures1, the impact will occur across many industries and affect many job tasks. The first breaks down jobs tasks by industry, and identifies banking, insurance, and software as industries that will be strongly impacted.

\n\n\n\n1 Both figures project a lower proportion of job tasks that AI can perform than the studies cited above. This could be because the projections in the former focus on LLMs, whereas the cited studies reference all forms of generative AI and other types of automation.","metadata":{}},{"cell_type":"markdown","source":"Figure 1: Large Language Model (LLM) Impact Across Industries (World Economic Forum, 2023)\n","metadata":{}},{"cell_type":"markdown","source":"And this figure, from the Goldman Sachs Report, is more task-focused, and highlights numerous job activities, such as administrative, legal, and financial functions.","metadata":{}},{"cell_type":"markdown","source":"Figure 2: Current Work Tasks That Could be Automated by AI in the\nUS: (Goldman Sachs, 2023)\n","metadata":{}},{"cell_type":"markdown","source":"In my opinion, and as perhaps indicated by the wide variety of industries and tasks cited above, generative AI's potential to complete work tasks is not fully understood, and is probably greater than we currently think.2 For one, employers are just beginning to apply the technology, so no projection can be taken as a certainty. Additionally, new Generative AI capabilities are being discovered or added very rapidly. In other words, we have not identified all of the tasks it can already perform, and there are others that perhaps it cannot handle as of this writing, but will be able to in the near future.\n

\n2 Relevantly, the authors of the McKinsey and Company study originally projected that 50% of jobs tasks could be replaced by generative AI and related technologies, but later revised this estimate upward to as high as 70% after further review.\n\n","metadata":{}},{"cell_type":"markdown","source":"## B.\tExperts Do Not Agree: How Will Generative AI Affect Employment Availability?\n\nThe title of the New York Times article \"The A.I. Revolution Will Change Work. Nobody Agrees How\" (Kessler, 2023) succinctly summarizes the landscape of opinions around generative AI's potential impact on jobs. The article notes that \n\"the equivalent of 300 million full-time jobs could be affected by A. I.\", but that \"affected\" in this context could mean \"made better, made worse, disappeared, doubled.” \n

\nFor instance, many suggest that the technology will create at least as many jobs as it replaces (e.g., Gaskrell, 2022). Proponents of this notion point to past technology-induced disruptions, noting that new technologies require new skillsets and therefore new jobs, and also that generative AI will lead to increased productivity and profitability, and thus more hiring.3

\n

\nHowever, given both the nature and volume of tasks generative AI can perform, I would caution against sanguinely concluding that \"no net job loss\" is the guaranteed or even most likely outcome. As already noted, the technology will impact an estimated hundreds of millions of jobs and 70% of job tasks--hardly a trivial volume. Additionally, no past technology has ever been able execute intellectual and creative tasks to the degree that generative AI can, even in this early stage of its development. One author identifies more than 70 such tasks it can handle as of right now including: software code, music, blog post, and email generation; software code review, completion, and bug fixing; streamlined drug discovery and development; educational course design; fashion design; and many others. Other technologies may have assisted workers in such tasks, but they have not been able to actually execute them. (Dilmegani, 2023).\n

\nOne might argue that humans are better than generative AI at these tasks. However, one author, concluding that \"ChatGPT is going to replace a lot of knowledge workers\", notes that many publishers of content don't require top quality output -- they will \"put up with less pithy headlines if they score more Google juice\" and \"lower quality articles if they generate revenue quicker.\" (Gerwitz, 2023). And of course generative AI's capabilities will improve over time.\n

\n3 Many proponents of this idea concede that, even if generative AI creates as many jobs as it eliminates, existing jobs are likely to be lost before new jobs are created (Vallance, 2023).\n \n \n ","metadata":{}},{"cell_type":"markdown","source":"In sum, it's likely that generative AI's impacts on jobs will differ from those of past technologies. Possible outcomes include:

\n\n### 1.\t Generative AI's Widespread Adoption Will Lead to Job Loss\n\nFor the reasons noted above, generative AI's adoption may very well result in some or even widespread job loss and a reduced quantity of job openings. In the words of one observer, “it’s not a question of if but when and how many specific jobs will be replaced by AI in near future.” (Aftab, 2022). \n

\n\nIn fact, this change is already in progress. As of mid-April 2023, to some degree because of generative AI technologies infiltrating the workplace, the tech industry cut 5% more jobs than it did in all of 2022 (Cohen, 2023), and companies such as IBM have announced that because of generative AI, they now allow attrition without replacing new jobs (Kelly, 2023). \n\n

\n\nHow many jobs could be lost? It's too early to know, but there are projections. For instance, a survey of global employers found that they expect AI will eliminate 83 million jobs by 2027, with 69 million new jobs created--a net loss of 14 million jobs, with administrative and clerical jobs most affected (Kelly, 2023). The Goldman-Sachs report referenced earlier suggests that up to 25% of all jobs in the US and Europe can be replaced entirely by generative AI. And Chat-GPT itself predicts that it will replace 4.8 million jobs (Cerullo, 2023), and that the workers it may replace include customer service representatives, technical writers, translators, interpreters, copywriters, and data entry clerks.\n\n\n","metadata":{}},{"cell_type":"markdown","source":"### 2.\t Those Who Remain Employed Will Not Need to Work Full-Time to Complete Job Tasks\n\nThose who argue that AI will create as many jobs as it will replace usually fail to note that even it does, it won’t take as much time to complete job tasks as it does today. For instance, according to a report from Accenture, 40% of all working hours could be affected by LLMs such as Chat-GPT (WEF, 2023), via either automation or augmentation.\nTo cite a few of many examples: two MIT economists recently found that ChatGPT allows professional writers to complete many tasks in half the usual time (Daniel, 2023); and generative AI can produce code in seconds, and considerably reduce time across other software development lifecycle functions, including testing, deployment, and maintenance (Cohen, 2023).

\nReducing the time taken to complete tasks might lead to a variety of results. For example, in a case where it takes 40 hours a week to do a job that previously took 400, nine jobs might be lost, or the employer might find 360 hours of new work to assign. But it's easy to envision cases where employees can't be assigned new work because there isn't enough work that generative AI cannot do, and the employer wants to retain some employees for subject matter expertise or other reasons. Thus, companies may look to reduce the number of hours that existing employees work. Since workers today generally receive health care, retirement, and other benefits only if they work full-time, this disruption could devastate jobs as we know them today\n\n\n\n","metadata":{}},{"cell_type":"markdown","source":"### 3.\t Generative AI Will Reduce Drudgery, Rote Work, and Physical Pain.\nAutomation in general and generative AI are by nature designed to handle repetitive and rote workplace tasks that people find boring and even dangerous. Many use cases for this change exist, such as entering and analyzing data in spreadsheets, matching job applicants to job opportunities, customer segmentation in marketing, and more, including ones we cannot anticipate at the moment. Additionally, generative AI in combination with robotics technology will be able to execute physically demanding and dangerous tasks, such as warehouse work and bartending. (Virlanuta, 2023).

\nInterestingly, some who champion this potential change echo the language and thinking of the libertarian critics of paid labor discussed above. One observes that because of AI, “we can automate mundane and repetitive tasks, freeing up time and energy for more meaningful and creative work.\"4 Another notes that generative AI “offers a better path to understand each other, to see clearly the humanity behind the employee number and to focus on unlocking the potential of every employee… it makes us, in the workplace and elsewhere, finally free to be more human\"4 (Webb, 2022). Perhaps such comments portend a future where we expect as a given that work (paid or otherwise) will offer true opportunities for purpose, connectivity, and meaning.\n

\n4 Emphases added","metadata":{}},{"cell_type":"markdown","source":"# IV.\t Generative AI Drives Expectations. Expectations Drive Outcomes\nTo drive my case home, I will summarize the above points, in the hope that my conclusions will follow logically:\n\n

    \n
  1. Human beings, by nature, desire and seek freedom, connectivity, and opportunities for creativity.\n
  2. Paid work, by its nature, blocks realization of these fundamental needs.\n
  3. Workplace managers have \"hacked\" human nature to improve employee productivity, by engineering jobs in order to increase employees' perception of control over their work.\n
  4. In recent times, workers and citizens have come to value and expect more autonomy and personal freedom as ends in themselves, not just as a means leveraged to a limited degree to help jobs seem more comfortable or tolerable.\n
  5. Generative AI will eliminate many full-time jobs. It will also change jobs that remain so they are less boring and repetitive and take less time to complete.\n
\n\nIn Western democracies, leaders must act in accordance with constraints the real world imposes and (in various degrees depending on how well the particular democracy functions) with the demands and aspirations of constituents. Currently, leaders respond to these drivers by implementing (or at least promising to implement) policies promoting proliferation of paid 40-hour per week jobs. But now, a new world is emerging whereby these jobs will be less widely available. Additionally, people are beginning to expect more freedom from the constraints such jobs typically impose, and generative AI, as a technology that can deliver these freedoms, will accelerate these expectations.

\n\nFor these reasons, as its fulfillment becomes less realistic and desirable, the full employment ethic will become less of an unchallenged societal value. As expectations change, so do societal and economic outcomes (Hatcher, 2018). In the face of many unknowns, precisely predicting specific outcomes is challenging, but the trends I have identified in this essay point to:\n

    \n
  1. An economy whereby, relative to today, a lower percentage of the population will hold or seek full-time jobs. \n
  2. Full-time jobs, to the degree they remain, that provide more meaningful and creative work and greater employee control over their location, time, and nature. \n
  3. Workplaces where employees will be able to complete their work more quickly and thereby have more time to themselves, without a substantial sacrifice of pay and benefits.\n
  4. Policies allowing citizens to meet their basic physiological needs whether or not they work full-time for wages.\n
\n","metadata":{}},{"cell_type":"markdown","source":"# V.\t Conclusion\n\nContrary to many dire predictions, generative AI and its widespread adoption in the workplace and elsewhere will not necessarily bring about a dystopian society. Rather, generative AI’s rapidly evolving capabilities to alter the command-control structure of paid work dovetail with evolving changes in cultural expectations. As a result, the technology offers opportunities for progress towards more fulfilling lives that better align to our essential human nature.

\n\nBut this does not mean utopia is on the horizon--progress always comes with new challenges. For instance, with fewer people working at full-time jobs, we would need alternatives to salaries in ensuring people’s basic survival needs are met. And some will resist the changes I’ve described—precisely because they potentially increase human freedom. Most Westerners have had their daily activities practically laid out for them their whole lives, first by school and then by jobs. Thus, many will find a rapid change involving greater freedom over one’s life choices threatening or scary. Addressing these challenges is beyond the scope of this essay, though I will note that they are not insurmountable (Lowrey, 2023).\n

\n\nI have tried to demonstrate Generative AI potentially can lead to a world offering more individual freedom and choice, but whether and the degree to which that happens is also a matter of choices—those we make as political and business leaders, constituents, and workers. I have identified characteristics and trends that I believe will influence those choices, but no one has a crystal ball and there are no guarantees.\n

\nIn my opinion, in the current day, most of us are expected to act as servants of economic growth and strength. Hopefully, our choices will lead to progress toward a reversal of this dynamic—to a world where technology, the workplace, and the larger economy serve humanity, not the other way around.","metadata":{}},{"cell_type":"markdown","source":"# References\n\nAftab, Nadeem (2022): Topmost Three Dangers of Artificial Intelligence.. Medium. Accessed Accessed June 21, 2023 https://medium.com/illumination/topmost-three-dangers-of-artificial-intelligence-58b8d12599f4 \n

\nBegley, Dave (2023): AI is far more deadly than nukes, The Truth Hurts News Share. Accessed June 19,2023, https://davebegleypoet.substack.com/p/ai-is-far-more-deadly-than-nukes\n

\nByrnes, Dave (2023): An AI bot has passed the bar exam. What happens next? Courthouse News Service. Accessed June 29, 2023, \nhttps://www.courthousenews.com/an-ai-bot-has-passed-the-bar-exam-what-happens-next\n

\nCarbonEdge (2023): The End of Mundane Tasks: Can AI Let You Focus on What Really Matters? Accessed June 27, 2023,https://www.carbonedge.com/insideedge-2/digital-transformation/future-of-work-ai\n

\nCerullo, Megan (2023): Here's how many U.S. workers ChatGPT says it could replace.CBS News. Accessed June 21, 2023 https://www.cbsnews.com/news/chatgpt-artificial-intelligence-jobs/\n

\nChomsky, Noam (1987): Language and Freedom. In The Chomsky Reader, edited by Noam Chomsky and James Peck, pp.139-155,\n

\nCohen, Mikaela (2023): These are the tech jobs most threatened by ChatGPT and A.I.. CNBC. Accessed Accessed June 21, 2023 https://www.cnbc.com/2023/04/15/these-are-the-tech-jobs-most-threatened-by-chatgpt-and-ai.html\n

\nDaniel, Will (2023): The A.I. boom will boost corporate profits 30% or more over the next decade, according to a Goldman Sachs senior strategist. Fortune. Accessed June 27, 2023. https://fortune.com/2023/05/18/artificial-intelligence-stock-market-impact-goldman-sachs-profits-30-percent-next-decade/\n

\nDilmegani, Cem (2023): Top 70+ Generative AI Applications / Use Cases in 2023. \nAIMultiple. Accessed July 14, 2024. https://research.aimultiple.com/generative-ai-applications/\n

\nFalk, Thomas M. (2021): Wage Slaves. In Encyclopedia of Critical Whiteness Studies in Education, edited by Zachary A. Casey, pp. 686-693.\n

\nGaskrell, Adi (2022): AI Creates Job Disruption But Not Job Destruction.Forbes. Accessed June 20, 2023\nhttps://www.forbes.com/sites/adigaskell/2022/01/18/ai-creates-job-disruption-but-not-job-destruction/?sh=7d245be53b3e\n

\nGerwitz, David (2023): Just how big is this new generative AI? Think internet-level disruption.ZDNet. Accessed July 14, 2023. https://www.zdnet.com/article/just-how-big-is-this-new-generative-ai-think-internet-level-disruption/\n

\nGoldman Sachs (2023): The Potentially Large Effects of Artificial Intelligence on\nEconomic Growth. Accessed June 20, 2023 https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html\n

\nHatcher, Michael (2018): Do Expectations Drive Big Events?Institute for Global Prosperity. Accessed June 28th, 2023. https://www.rebuildingmacroeconomics.ac.uk/post/do-expectations-drive-big-events\n

\nKelly, Jack (2023): Artificial Intelligence Is Already Causing Disruption And Job Losses At IBM And Chegg.Forbes. Accessed June 20, 2023 https://www.forbes.com/sites/jackkelly/2023/05/02/artificial-intelligence-is-already-causing-disruption-and-job-losses-at-ibm-and-chegg/?sh=2e0f362d75a4\n

\nKessler, Sarah (2023): The A.I. Revolution Will Change Work. Nobody Agrees How.. New York Times. Accessed June 20,2023, https://www.nytimes.com/2023/06/10/business/ai-jobs-work.html \n

\nKoning, Peter (2020): How Much Autonomy Should Teams get from Their Agile Leader?. Scrum.org. Accessed June 20,2023, https://www.scrum.org/resources/blog/how-much-autonomy-should-teams-get-their-agile-leader\n

\nLowrey, Annie (2023): Before AI Takes Over, Make Plans to Give Everyone Money. The Atlantic. Accessed June 28,2023, https://www.theatlantic.com/ideas/archive/2023/05/ai-job-losses-policy-support-universal-basic-income/674071\n

\nMarr, Bernard (2023): AI Anxiety: How These 20 Jobs Will Be Transformed By Generative Artificial Intelligence.. Forbes. Accessed June 20, 2023 https://www.forbes.com/sites/bernardmarr/2023/06/05/ai-anxiety-how-these-20-jobs-will-be-transformed-by-generative-artificial-intelligence/?sh=36de8dc34b45\n

\nMckinsey & Company (2023a): What is Generative AI? accessed June 19,2023, https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-generative-ai\n

\nMckinsey & Company (2023b): The economic potential of generative AI: The next productivity frontier. Accessed June 20,2023, https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier#key-insights\n

\nMoss, Jennifer (2022): The Pandemic Changed Us. Now Companies Have to Change Too. Harvard Business Review. Accessed June 20, 2023, https://hbr.org/2022/07/the-pandemic-changed-us-now-companies-have-to-change-too\n

\nPolin, Robert (2011): Back to Full Employment, Boston Review accessed June 19,2023 https://www.bostonreview.net/forum/back-full-employment\n

\nRyan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78. Accessed June 20, 2023 https://selfdeterminationtheory.org/SDT/documents/2000_RyanDeci_SDT.pdf\n

\nSpector, Paul. E. (1986): Perceived Control by Employees: A Meta-Analysis of Studies Concerning Autonomy and Participation at Work. Human Relations, 39(11), 1005–1016. Accessed June 20, 2023 https://doi.org/10.1177/001872678603901104\n

\nVallance, Chris (2023): AI could replace equivalent of 300 million jobs - report. BBC. Accessed June 21, 2023. https://www.bbc.com/news/technology-65102150 \n

\nVirlanuta, Claudia (2023): How to Automate Your Boring Tasks with Machine Learning. Edlitera. Accessed June 27, 2023,https://www.edlitera.com/en/blog/posts/automate-boring-tasks-with-ml.\n

\nWebb, Geoff (2022): The Final Freedom: The Role Of Artificial Intelligence In The Workplace. Forbes. Accessed June 27, 2023, https://www.forbes.com/sites/forbestechcouncil/2022/12/09/the-final-freedom-the-role-of-artificial-intelligence-in-the-workplace/?sh=644f71e95e69\n

\nWorld Economic Forum (2023): These are the jobs most likely to be lost – and created – because of AI. Accessed June 27, 2023. https://www.weforum.org/agenda/2023/05/jobs-lost-created-ai-gpt/\n \n\n \n \n\n\n","metadata":{}}]}