Three Years in Four Weeks: How Enterprise AI Rewrites Employment

In late February 2026, Perplexity AI quietly published a blog post with a claim that should have set off alarms in every corporate office from London to Los Angeles. The company's new product, Computer for Enterprise, had been deployed internally as a Slack integration, with every employee in the same channel. After processing more than 16,000 queries in four weeks, the system had, by Perplexity's own estimation, completed the equivalent of 3.25 years of human work and saved the company $1.6 million in labour costs. The benchmarks used to measure this output came from institutions including McKinsey, Harvard, MIT, and Boston Consulting Group.

Let that settle for a moment. Not 3.25 years spread across thousands of workers performing marginal speed improvements. The claim is that a single AI platform, running cloud-based workflows across roughly 20 frontier models, replaced years of the kind of cognitive labour that knowledge workers perform every day: querying databases, compiling reports, synthesising research, drafting analyses. The tasks that fill the calendars of financial analysts, marketing strategists, management consultants, and corporate researchers everywhere.

Perplexity's CEO, Aravind Srinivas, framed the ambition with characteristic directness. “What we are going to try to do is help businesses run as autonomously as possible,” he said. On the question of AI displacing jobs, he offered a response that managed to be both provocative and revealing: “The reality is most people don't enjoy their jobs.” His suggestion was that displacement could free people to pursue entrepreneurship and more fulfilling work. It is, to put it mildly, an incomplete answer to a question affecting hundreds of millions of workers worldwide.

The Machine That Writes the Queries

To understand why Perplexity's claims matter, you need to understand what Computer for Enterprise actually does. It is not a chatbot. It is not a search engine with a conversational veneer. It is an orchestration platform that routes tasks across approximately 20 AI models from multiple providers, including Anthropic's Claude Opus 4.6 as its primary reasoning engine, Google's Gemini for deep research, OpenAI's GPT-5.2, and xAI's Grok. Each session runs inside its own isolated Firecracker virtual machine, ensuring data separation between users.

The platform connects natively to the software stack that modern enterprises already run: Snowflake, Salesforce, HubSpot, Slack, Notion, GitHub, Gmail, Outlook, and more than 400 other applications through its connector ecosystem. Administrators can install custom connectors via the Model Context Protocol. The system includes workflow templates for legal contract review, finance audit support, sales call preparation, and customer support ticket triage.

Here is the critical capability: Computer for Enterprise does not merely answer questions. It writes the database queries, executes them, and returns structured results. A financial analyst can ask for revenue broken down by vertical from Snowflake, and the system will compose the SQL, run it against the data warehouse, and present the findings. A sales team can simultaneously pull CRM data and competitive context. The AI handles the translation from natural language intent to technical execution and back again, collapsing what might take a human analyst hours into seconds.

Srinivas described the underlying philosophy on the social media platform X: “When AIs can orchestrate a file system with CLI tools plus a browser, AI essentially becomes the Computer, running things on the cloud as you sleep.” He drew a distinction between traditional operating systems and what Perplexity is building: “A traditional operating system takes instructions; an AI operating system takes objectives.”

The enterprise offering comes wrapped in the security apparatus that corporate procurement teams demand: SOC 2 Type II compliance, SAML single sign-on, audit logs, sandboxed query execution, and GDPR and HIPAA compliance. Pricing runs at $325 per user per month for the Enterprise Max tier, or $40 per user per month for Enterprise Pro. Perplexity's annualised revenue reached approximately $148 million by mid-2025, with internal projections targeting $656 million by the end of 2026.

The company is candid about limitations. Factual hallucinations occur, particularly on niche topics or very recent events. The system occasionally generates broken URLs. External communications, whether emails or published content, should always be reviewed by a human before distribution. But the trajectory is clear, and the implications are staggering.

The Scale of What Could Be Lost

The question that Perplexity's announcement forces into the open is not whether AI can perform knowledge work. That debate ended sometime around mid-2024, when large language models began consistently demonstrating competence at research synthesis, data analysis, report writing, and code generation. The question now is what happens to the people who currently do this work for a living.

The numbers are sobering. According to Goldman Sachs research, generative AI could automate tasks equivalent to 300 million full-time jobs worldwide, with 26 per cent of office roles and 20 per cent of customer service positions highly exposed. In the United States alone, Goldman Sachs estimates that AI automation will ultimately displace roughly six to seven per cent of the workforce, equivalent to approximately 11 million workers. The World Economic Forum's Future of Jobs Report 2025, drawing on perspectives from more than 1,000 leading global employers representing over 14 million workers, projects that 92 million roles will be displaced by 2030, though it forecasts 170 million new roles emerging for a net gain of 78 million jobs.

McKinsey's analysis adds another dimension. The consultancy estimated that today's technology could, in theory, automate approximately 57 per cent of current U.S. work hours. That figure does not mean 57 per cent of jobs will vanish. It means that across the entire working population, just over half of the hours worked involve tasks that a sufficiently deployed AI system could handle. McKinsey projects that 30 per cent of U.S. work hours could be automated by 2030, accelerated by generative AI's capabilities.

The disruption is already visible in employment data. There were 77,999 AI-attributed tech job losses in the first six months of 2025 alone. Employment in the computer systems design and related services sector declined five per cent since ChatGPT's release. Entry-level job postings dropped 15 per cent year over year. Employment among software developers aged 22 to 25 fell 20 per cent compared to their late 2022 peak. According to research from the Dallas Federal Reserve, AI is simultaneously aiding existing workers and replacing others, with the wage data suggesting a complex and uneven transformation.

Certain roles face particularly acute risk. Data entry positions carry a 95 per cent automation risk. Customer service representatives face 80 per cent risk, because most inquiries are answerable from a knowledge base. Paralegals face an 80 per cent risk of automation by 2026, and legal researchers face a 65 per cent risk by 2027. An estimated 200,000 jobs are expected to be cut from Wall Street banks over the next three to five years, and as much as 54 per cent of banking jobs have high potential for AI automation. SSRN projections estimate that 7.5 million data entry and administrative jobs could be eliminated by 2027.

Seventy-five per cent of knowledge workers are already using AI tools at work, and nearly half started within the last six months. They report 66 per cent productivity improvements. But the question nobody wants to confront directly is this: if each worker becomes 66 per cent more productive, how many fewer workers does an organisation actually need?

The Cautionary Tale Already Playing Out

The corporate world is not waiting for the research to settle before acting. The global technology sector eliminated nearly 60,000 jobs in less than three months of 2026, according to layoff tracker TrueUp, which recorded 171 separate events affecting 59,121 workers since January. That pace, averaging 704 jobs lost per day, is running ahead of 2025, when 245,953 workers were let go across the full year. If it holds, total cuts could reach 265,000 by December. A Resume.org survey of 1,000 U.S. hiring managers found that 55 per cent expect layoffs at their companies in 2026, and 44 per cent identified AI as a primary driver.

Some of the largest names in technology are leading the charge. Amazon confirmed 16,000 corporate job cuts in 2026 despite reporting record revenue of $716.9 billion the previous year, framing the reductions as a push to flatten management layers. Some of those roles are not being backfilled with humans; they are being backfilled with software. Block, the payments company formerly known as Square, slashed 4,000 roles in early 2026, nearly 40 per cent of its entire workforce. Ingka Group, the largest IKEA retailer, announced 800 office role cuts in March.

Perhaps the most instructive example comes from Klarna, the Swedish fintech company. In 2024, Klarna deployed an AI assistant that handled the equivalent workload of 700 full-time customer service employees. The company's headcount fell from approximately 7,000 in 2022 to roughly 3,000, and CEO Sebastian Siemiatkowski publicly championed the results. But the strategy backfired. Customer complaints increased, satisfaction ratings dropped, and internal reviews revealed that AI systems lacked empathy and could not handle nuanced problem-solving. By early 2025, Siemiatkowski acknowledged that the company had overestimated AI's capabilities, stating bluntly: “We went too far.” Klarna began rehiring human customer service staff, specifically targeting students, rural populations, and dedicated product users.

Klarna's reversal is a cautionary tale that speaks directly to Acemoglu's warnings about “so-so automation.” The financial savings looked impressive on a spreadsheet, but the technology degraded the quality of the service it was supposed to improve. The question for every organisation evaluating tools like Perplexity's Computer for Enterprise is whether the same pattern will repeat across other domains: impressive benchmarks followed by the slow realisation that human judgement, context, and empathy were doing more work than anyone appreciated until they were gone.

The Uncomfortable History of “New Jobs Will Appear”

Every wave of technological disruption produces two competing narratives. The optimists point to history: the Industrial Revolution destroyed agricultural and artisan livelihoods but created factory work. The IT revolution eliminated typing pools and filing clerks but created entire industries around software, networking, and digital services. The pessimists counter that this time is different, that the pace and breadth of AI's capabilities outstrip anything that came before.

History offers both comfort and caution. During the first Industrial Revolution, the Luddites famously destroyed the mechanised looms that threatened their livelihoods in industrial Britain. Their fears were not irrational. While new manufacturing jobs eventually emerged, the transition period was brutal. Research from economic historians shows that average real wages in England stagnated for decades even as productivity rose. Eventually, wage growth caught up to and then surpassed productivity growth, but only after substantial policy reforms including labour protections and education acts.

The Second Industrial Revolution followed a similar pattern. Automation technologies increased the efficiency and scope of mechanised production, requiring fewer operators but more engineers, managers, and other new occupations. As automation created fewer middle-skill jobs than it made obsolete, the result was a hollowing out of the skill distribution in manufacturing, a pattern that persists to this day.

The robotics wave of the 1970s and 1980s displaced approximately 1.2 million manufacturing jobs globally by 1990. In the United States alone, robot-induced automation displaced 300,000 factory workers in the automotive sector. New jobs did eventually appear, but they required different skills, existed in different locations, and often paid different wages.

McKinsey's historical analysis offers a striking statistic: 60 per cent of today's U.S. workforce is employed in occupations that simply did not exist in 1940. That is genuinely encouraging. But it also means that 60 per cent of today's workers are in roles that their grandparents could not have trained for, because the jobs had not yet been invented. The lag between destruction and creation is where the human cost concentrates.

What makes the AI wave qualitatively different from previous automation episodes is its target. Earlier forms of automation primarily replaced physical labour and routine cognitive tasks: drilling, sewing, sorting files, calculating spreadsheets. AI encroaches on non-routine cognitive domains once thought uniquely human, including recognising images, drafting emails, drawing illustrations, synthesising research, and making complex judgements. The Bipartisan Policy Center in Washington notes that AI is different because it can automate many tasks that do not follow an explicit set of rules and are instead learned through experience and intuition.

The pace compounds the challenge. Previous technological transitions unfolded over generations, allowing social institutions to adapt. The shift from agricultural to industrial employment in the United States took roughly a century. The transition from manufacturing to services took several decades. AI capabilities are advancing on a timeline measured in months. Goldman Sachs models show that each one percentage point productivity gain from technology raises unemployment by approximately 0.3 percentage points in the short run, though this effect historically fades within two years.

Who Gets Hurt First

The distributional question matters enormously. The World Economic Forum's net positive headline of 78 million new jobs conceals what the organisation itself acknowledges is a profound distributional challenge: the jobs being destroyed and the jobs being created are not the same jobs, do not require the same skills, do not pay the same wages, and are not located in the same geographies.

Entry-level and young workers are bearing the brunt. AI can replicate codified knowledge but not tacit knowledge, the experiential understanding that comes from years of practice. This means AI may substitute for entry-level workers while augmenting the efforts of experienced professionals. Fourteen per cent of all workers report having already been displaced by AI, with the rate higher among younger and mid-career workers in technology and creative fields. Unemployment among 20 to 30 year olds in tech-exposed occupations has risen by almost three percentage points since the start of 2025, according to Goldman Sachs data, notably higher than for their same-aged counterparts in other trades.

There is also a significant gender dimension. In the United States, 79 per cent of employed women work in jobs that are at high risk of automation, compared to 58 per cent of men. That translates to 58.87 million women versus 48.62 million men occupying positions highly exposed to AI automation.

White-collar workers in industries such as financial services and media now express higher levels of concern about automation (67 per cent) than their counterparts in blue-collar sectors, including transportation (60 per cent) and retail (59 per cent). The traditional assumption that automation primarily threatens manual and routine work has been comprehensively upended. AI poses a risk of eliminating 10 to 20 per cent of entry-level white-collar jobs within the next one to five years.

The irony is sharp. Knowledge workers spent decades insulating themselves from automation risk by acquiring education, developing analytical skills, and moving into roles that required judgement and communication. Now the very capabilities they cultivated, research synthesis, data analysis, report writing, pattern recognition, are precisely what large language models do best.

The Productivity Paradox

Not all economists agree on the magnitude of the disruption. Daron Acemoglu, the Nobel Prize-winning economist and Institute Professor at MIT, offers one of the most rigorously evidence-based counterpoints to the prevailing AI hype. Despite predictions from some quarters that AI will dramatically boost GDP growth, Acemoglu expects it to increase U.S. GDP by just 1.1 to 1.6 per cent over the next decade, with a roughly 0.05 per cent annual gain in productivity. He believes current AI tools are likely to impact only about five per cent of jobs.

Acemoglu's central concern is what he terms “so-so automation,” technologies that replace jobs without meaningfully boosting productivity or human welfare. “When hype takes over, companies start automating everything, including tasks that shouldn't be automated,” he has warned. “You end up with no productivity gains, damaged businesses, and people losing jobs without new opportunities being created.” Think of self-checkout kiosks that are slower and more frustrating than human cashiers, or automated customer service menus that leave callers trapped in loops of increasingly desperate button-pressing.

His prescription is pointed: “We're using it too much for automation and not enough for providing expertise and information to workers.” He draws a crucial distinction between AI that provides new information to a biotechnologist, helping them become more effective, and AI that replaces a customer service worker with an automated system. The former creates value; the latter merely transfers costs from employer to consumer.

Acemoglu acknowledges that AI will transform many occupations but remains sceptical of elimination claims: “I don't expect any occupation that we have today to have been eliminated in five or 10 years' time. We're still going to have journalists, we're still going to have financial analysts, we're still going to have HR employees.” What will change, he argues, is the task composition within those roles, with AI handling data summary, visual matching, and pattern recognition while humans focus on judgement, creativity, and interpersonal skills.

Gartner's projections align with this more measured view, predicting that AI's impact on global jobs will be neutral through 2026, and that by 2028, AI will create more jobs than it destroys. But neutral aggregate impact can still mask severe disruption for specific communities, industries, and demographics.

The Scramble to Adapt

Organisations are responding with a mixture of enthusiasm and anxiety. According to the World Economic Forum, 41 per cent of employers globally plan to use AI to reduce headcount, while simultaneously 77 per cent aim to upskill their staff for working alongside AI, and 47 per cent plan to move affected employees into different roles internally. About one in six employers expect AI to reduce headcount in 2026 specifically.

The skills gap is already the most significant barrier to business transformation, with nearly 40 per cent of skills required on the job set to change and 63 per cent of employers citing it as their key challenge. The number of workers in occupations where AI fluency is explicitly required has risen from around one million in 2023 to approximately seven million in 2025, according to McKinsey data. Across McKinsey's most recent global survey, 94 per cent of employees and 99 per cent of C-suite executives report personal use of generative AI.

Companies are pursuing several adaptation strategies simultaneously. Some are integrating AI with their proprietary data via retrieval-augmented generation or fine-tuning, creating what Goldman Sachs describes as expert AI systems with advanced capabilities and industry-specific knowledge. Others are restructuring roles around human-AI collaboration, keeping the human in the loop for judgement calls, client relationships, and strategic decisions while delegating research, analysis, and first-draft creation to AI systems. According to a PwC survey of 300 senior executives conducted in May 2025, 88 per cent said their team or business function plans to increase AI-related budgets in the next twelve months due to agentic AI, while 79 per cent reported that AI agents are already being adopted in their companies.

The retraining challenge, however, is formidable. The half-life of professional skills is collapsing faster than any training programme can keep pace with. A displaced worker who enrols in an eighteen-month data analytics programme may find that entry-level positions in that field have already been automated by graduation. Nobel laureate Angus Deaton has noted that economists were naively optimistic about the effectiveness of trade adjustment assistance, including worker retraining programmes, for those hurt by previous economic shifts. The track record of large-scale retraining initiatives is, at best, mixed.

PwC's own research underscores a deeper challenge: technology delivers only about 20 per cent of an initiative's value. The other 80 per cent comes from redesigning work so that AI agents can handle routine tasks and people can focus on what truly drives impact. That redesign requires not just new software licences but fundamental rethinking of roles, workflows, and organisational structures. It is the kind of transformation that most companies talk about but few execute well.

The Policy Vacuum

The policy conversation is struggling to keep pace with the technology. In early 2026, U.K. Minister for Investment Lord Jason Stockwood told the Financial Times that the government is weighing the introduction of a universal basic income to support workers in industries where AI threatens displacement. “Undoubtedly we're going to have to think really carefully about how we soft-land those industries that go away,” he said, “so some sort of UBI, some sort of lifelong learning mechanism as well so people can retrain.” He has also floated the idea of technology companies being taxed to fund such payments.

The UBI discussion has shifted from theoretical curiosity to practical policy consideration. Ioana Marinescu, an economist at the University of Pennsylvania, has argued that UBI could be a pragmatic solution to AI-driven job displacement, particularly given the uncertainties around how many people will lose their jobs and for how long. For people without prior employment history, especially younger workers entering the labour market for the first time, unemployment insurance benefits are not guaranteed, making unconditional UBI payments a potentially effective safety net.

The idea has precedent. According to the Stanford Basic Income Lab, 163 programmes piloting basic income, including 41 active programmes, have been run in the United States alone. Ireland's Basic Income for the Arts programme, which began as a three-year pilot, will become permanent in 2026, allowing creative workers to pursue their craft without needing supplementary employment.

Researchers at the London School of Economics argue that UBI's successful implementation depends on sustainable funding mechanisms, investment in education, and attention to social and psychological dimensions, not only economic and labour market outcomes. The question of funding remains contentious. In 2017, Bill Gates proposed taxing robots, suggesting that companies replacing human workers with automation should pay taxes at levels comparable to the people they displace. The concept of an AI automation tax is gaining traction as a revenue source where automation's economic benefits help support those most affected by the transition.

Morgan Stanley noted in a report in early 2026 that AI-related job cuts are hitting Britain the hardest, with eight per cent net job losses over the preceding twelve months. The United States currently has no comprehensive labour transition strategy, no reskilling infrastructure capable of operating at the required speed, and no serious public conversation about income decoupled from employment.

Some analysts advocate for integrated approaches: AI-enabled personalised retraining pathways, job matching to emerging sectors, and combining UBI with reskilling initiatives, education grants, and healthcare services. Policymakers are urged to prioritise pilot programmes that integrate income support with workforce development, leveraging AI itself to optimise distribution and measure impact.

The Tension That Will Not Resolve

The fundamental tension at the heart of this story has no clean resolution. Perplexity's Computer for Enterprise represents a genuine productivity breakthrough. If knowledge workers can accomplish in seconds what previously took hours, the economic potential is enormous. Organisations that adopt these tools will move faster, spend less on routine analysis, and free their best people to focus on the creative and strategic work that AI still handles poorly.

But the maths of productivity improvement and the maths of employment are not the same calculation. When Srinivas says he wants to help businesses run as autonomously as possible, he is describing a world with fewer employees. When Perplexity's internal study shows 3.25 years of work completed in four weeks, it is demonstrating that the same output can be achieved with a fraction of the human input. When 75 per cent of knowledge workers report using AI and seeing 66 per cent productivity gains, the logical endpoint is that organisations need significantly fewer knowledge workers to produce the same volume of output.

The World Economic Forum projects a net positive outcome globally, with new job categories emerging to replace those that disappear. History suggests this is likely correct over sufficiently long time horizons. But the transition period, the years between when old jobs vanish and new ones coalesce, is where lives are disrupted, careers are derailed, mortgages go unpaid, and communities fracture. Klarna's experience is a reminder that even the companies most aggressively pursuing AI-driven efficiency can discover, too late, that they have optimised away something essential.

Acemoglu urges a more deliberate approach: deploying AI to augment human capabilities rather than simply replacing human workers, celebrating what he calls “the places where AI is better than humans, and the places where humans are better than AI.” Given the mixed evidence on benefits and drawbacks, he and his colleagues argue that it may be best to adopt AI more slowly than market fundamentalists might prefer.

That counsel of patience, however, runs headlong into competitive reality. No company can afford to ignore a technology that promises to compress years of work into weeks, not when their competitors are already adopting it. The individual incentive to automate is overwhelming, even if the collective consequence is displacement on a scale that existing social safety nets were never designed to absorb.

Srinivas outlined an AI evolution on LinkedIn: “2023: Using AI to research. 2024: Super prompting galore. 2025: AI remembers you. 2026: Agents are useful (and not just to vibe coders).” He added that intelligence is no longer the bottleneck; what matters now is knowing which model to call, what context to surface, and when to act versus ask a follow-up question.

For the millions of knowledge workers whose professional identity is built on exactly those skills, research, analysis, synthesis, and communication, the message is unsettling. The tools that made their expertise valuable are now embedded in software that costs $325 per month and never sleeps. The question is not whether the transformation will happen. It is whether societies will manage the transition with anything approaching the speed, scale, and seriousness that the moment demands. Based on every previous technological transition in recorded history, the honest answer is: probably not fast enough.

References

  1. Perplexity AI, “Computer for Enterprise,” Perplexity Blog, March 2026. https://www.perplexity.ai/hub/blog/computer-for-enterprise
  2. PYMNTS, “Perplexity's Computer for Enterprise Completed 3.25 Years of Work in Four Weeks,” PYMNTS.com, March 2026. https://www.pymnts.com/news/artificial-intelligence/2026/perplexity-computer-enterprise-completed-3-years-work-4-weeks/
  3. Fortune, “Perplexity CEO explains Computer, its OpenClaw-like AI agent tool for non-experts,” Fortune, February 2026. https://fortune.com/2026/02/26/perplexity-ceo-aravind-srinivas-computer-openclaw-ai-agent/
  4. Fortune, “Perplexity CEO Aravind Srinivas: AI layoffs aren't so bad as 'most people don't enjoy their jobs',” Fortune, March 2026. https://fortune.com/2026/03/24/perplexity-ceo-ai-layoffs-not-bad-people-hate-jobs-entrepreneurship/
  5. VentureBeat, “Perplexity takes its 'Computer' AI agent into the enterprise, taking aim at Microsoft and Salesforce,” VentureBeat, March 2026. https://venturebeat.com/technology/perplexity-takes-its-computer-ai-agent-into-the-enterprise-taking-aim-at
  6. The Register, “Perplexity: Everything is Computer,” The Register, March 2026. https://www.theregister.com/2026/03/12/perplexity_extends_cloud_computer_to_enterprise/
  7. Goldman Sachs, “How Will AI Affect the Global Workforce?,” Goldman Sachs Insights, 2024. https://www.goldmansachs.com/insights/articles/how-will-ai-affect-the-global-workforce
  8. Goldman Sachs, “What to expect from AI in 2025: hybrid workers, robotics, expert models,” Goldman Sachs Insights, 2025. https://www.goldmansachs.com/insights/articles/what-to-expect-from-ai-in-2025-hybrid-workers-robotics-expert-models
  9. World Economic Forum, “Future of Jobs Report 2025,” WEF, January 2025. https://www.weforum.org/publications/the-future-of-jobs-report-2025/
  10. McKinsey, “AI in the workplace: A report for 2025,” McKinsey & Company, 2025. https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work
  11. McKinsey, “Five lessons from history on AI, automation, and employment,” McKinsey & Company, 2024. https://www.mckinsey.com/featured-insights/future-of-work/five-lessons-from-history-on-ai-automation-and-employment
  12. Dallas Federal Reserve, “AI is simultaneously aiding and replacing workers, wage data suggest,” Federal Reserve Bank of Dallas, February 2026. https://www.dallasfed.org/research/economics/2026/0224
  13. MIT Technology Review, “A Nobel laureate on the economics of artificial intelligence,” MIT Technology Review, February 2025. https://www.technologyreview.com/2025/02/25/1111207/a-nobel-laureate-on-the-economics-of-artificial-intelligence/
  14. MIT Economics, “Daron Acemoglu: What do we know about the economics of AI?,” MIT Economics, 2025. https://economics.mit.edu/news/daron-acemoglu-what-do-we-know-about-economics-ai
  15. Bipartisan Policy Center, “What Past Waves of Automation Can Teach Us About AI,” BPC, 2024. https://bipartisanpolicy.org/article/what-past-waves-of-automation-can-teach-us-about-ai/
  16. Brookings Institution, “Measuring US workers' capacity to adapt to AI-driven job displacement,” Brookings, 2025. https://www.brookings.edu/articles/measuring-us-workers-capacity-to-adapt-to-ai-driven-job-displacement/
  17. Fortune, “UK minister calls for universal basic income to cushion the blow from AI-related job losses,” Fortune, February 2026. https://fortune.com/2026/02/01/elon-musk-optional-work-fantasy-universal-basic-income-uk-minister-jason-stockwood/
  18. LSE Business Review, “Universal basic income as a new social contract for the age of AI,” London School of Economics, April 2025. https://blogs.lse.ac.uk/businessreview/2025/04/29/universal-basic-income-as-a-new-social-contract-for-the-age-of-ai-1/
  19. The Hill, “AI Revolution: The Case for Universal Basic Income,” The Hill, 2025. https://thehill.com/opinion/finance/5713876-ai-displacement-and-ubi/
  20. ALM Corp, “AI Job Displacement Statistics 2025-2030: 60+ Data Points,” ALM Corp, 2025. https://almcorp.com/blog/ai-job-displacement-statistics/
  21. DemandSage, “77 AI Job Replacement Statistics 2026 (New Data),” DemandSage, 2026. https://www.demandsage.com/ai-job-replacement-stats/
  22. Perplexity AI, “Everything is Computer,” Perplexity Blog, February 2026. https://www.perplexity.ai/hub/blog/everything-is-computer
  23. Perplexity AI, “The Intelligent Business: Introducing Comet for Enterprise Pro,” Perplexity Blog, 2026. https://www.perplexity.ai/hub/blog/the-intelligent-business-introducing-comet-for-enterprise-pro
  24. Morgan Stanley, AI-related job displacement report, cited in AllWork.space, February 2026. https://allwork.space/2026/02/universal-basic-income-reenters-the-future-of-work-debate-as-ai-disrupts-jobs/
  25. IBTimes, “Tech Layoffs Surge to 59,000 in 2026 as Amazon, Meta and Block Cut Jobs Amid AI Shift,” International Business Times, March 2026. https://www.ibtimes.com/tech-layoffs-surge-59000-2026-amazon-meta-block-cut-jobs-amid-ai-shift-whos-next-3800066
  26. Tech.co, “Klarna Reverses AI Customer Service Replacement,” Tech.co, 2025. https://tech.co/news/klarna-reverses-ai-overhaul
  27. MLQ.ai, “Klarna CEO admits AI job cuts went too far,” MLQ.ai, 2025. https://mlq.ai/news/klarna-ceo-admits-aggressive-ai-job-cuts-went-too-far-starts-hiring-again-after-us-ipo/
  28. PwC, “AI Agent Survey,” PwC, 2025. https://www.pwc.com/us/en/tech-effect/ai-analytics/ai-agent-survey.html
  29. Stanford Basic Income Lab, UBI pilot programme data, Stanford University. https://basicincome.stanford.edu
  30. Daron Acemoglu, “The Simple Macroeconomics of AI,” MIT Economics Working Paper, 2024. https://economics.mit.edu/sites/default/files/2024-04/The%20Simple%20Macroeconomics%20of%20AI.pdf

Tim Green

Tim Green UK-based Systems Theorist & Independent Technology Writer

Tim explores the intersections of artificial intelligence, decentralised cognition, and posthuman ethics. His work, published at smarterarticles.co.uk, challenges dominant narratives of technological progress while proposing interdisciplinary frameworks for collective intelligence and digital stewardship.

His writing has been featured on Ground News and shared by independent researchers across both academic and technological communities.

ORCID: 0009-0002-0156-9795 Email: tim@smarterarticles.co.uk

Discuss...