---
title: "Artificial Intelligence"
description: "Artificial intelligence (AI) has quickly become one of history’s most revolutionary technological innovations, reshaping industries, attracting massive capital…"
url: https://www.independentpress.com/article/artificial-intelligence
date: 2026-08-07
categories: ["Artificial Intelligence","Economics","Social"]
---

# Artificial Intelligence

![The Creation of Adam Sistine Chapel](https://images.ctfassets.net/ewtdlsoyixc1/3MK1t3CUDMuMNkaZEH75ZI/10d45bdbfb485aa542fc2db212572689/The_Creation_of_Adam_Sistine_Chapel.jpg)

Artificial intelligence (AI) has quickly become one of history’s most revolutionary technological innovations, reshaping industries, attracting massive capital allocations, and dominating headlines. Governments and corporations alike are investing heavily, betting that AI will unlock new levels of automation, productivity, and economic growth. But with this surge in enthusiasm comes a growing concern: have we gone too far, too fast? 

The scale of investment is staggering. Technology giants like Amazon, Google, Meta, and Microsoft have spent hundreds of billions on AI infrastructure, driven as much by competitive pressure and fear of missing out as by strategic planning. Yet the return on this spending remains uncertain. The market is already showing signs of fatigue; jobs are being cut, critical water and energy infrastructure is buckling under the increasing demands of artificial intelligence, and concerns regarding AI’s impact on society are rising.

As costs continue to mount, the world is being forced to confront deeper, longer-term tradeoffs. If artificial intelligence continues to replace everyday social interactions, we risk losing something far more important than economic prosperity: we risk losing our own humanity. With the rapid, unabated advancement of AI, we are in danger of prioritizing technological pursuit over the wellbeing of our people and society. Though artificial intelligence has promised wide-ranging benefits, the upfront costs have escalated precipitously, outpacing expectations and raising concerns that we have entered a severe economic bubble where resources are being committed haphazardly without prudence or foresight. While AI holds immense economic potential, the current pace and scale of investment must now reckon with the compromises we have made in pursuit of technology at the expense of economic, social, and environmental sustainability.

**The Emergence of AI**

The origins of artificial intelligence can be traced back to the 20th century when mathematician and computer scientist [Alan Turing](https://www.britannica.com/biography/Alan-Turing) proposed the question “Can Machines Think?” in his landmark 1950 paper, [_Computing Machinery and Intelligence_](https://courses.cs.umbc.edu/471/papers/turing.pdf). Turing’s work on cryptography during World War II and his proposal of the [Turing Test](https://plato.stanford.edu/entries/turing-test/) helped lay the foundation for the transformative technology we know as AI. The Turing Test, a method for evaluating a machine's ability to exhibit intelligent behavior comparable to a human, provided a benchmark for testing machine intelligence and inspired the creation of early artificial intelligence models. 

Artificial intelligence continued to evolve in waves, first with symbolic AI and rule-based expert systems in the 1950s through the 1980s, as machines were programmed with specific rules, and then with the transition to machine learning systems from the late 1980s through the 2000s, when computers began learning from data rather than restrictive programming. AI development gained further momentum as key figures such as [Geoffrey Hinton](https://www.nobelprize.org/prizes/physics/2024/hinton/facts/) pioneered neural networks and backpropagation algorithms, [Yann LeCun](https://quantumzeitgeist.com/yann-lecun-the-french-ai-pioneer-behind-the-convolutional-neural-network/) developed convolutional neural networks improving image recognition technology, and [Stuart Russell](https://people.eecs.berkeley.edu/~russell/) advanced human-compatible AI alignment and safety frameworks. In the 21st century, [deep learning](https://www.ibm.com/think/topics/deep-learning)—fueled by increased computational power from computer chips and informational availability from data repositories and indexing engines like Google—pushed AI into a new era. Advances such as [Nvidia’s GPUs](https://www.nvidia.com/en-us/solutions/ai/)combined with the explosion of cloud computing platforms such as [Amazon Web Services (AWS)](https://aws.amazon.com/ai/machine-learning/), [Microsoft Azure](https://azure.microsoft.com/en-us), and [Google Cloud](https://cloud.google.com/free?utm_source=google&utm_medium=cpc&utm_campaign=Cloud-SS-DR-GCP-1713658-GCP-DR-NA-US-en-Google-BKWS-EXA-GeneralGCP&utm_content=c-Hybrid+%7C+BKWS+-+EXA+%7C+Txt-Generic+Cloud-Cloud+Generic-Cloud+Generic-6458750523&utm_term=google%20cloud&gclsrc=aw.ds&gad_source=1&gad_campaignid=23752515549&gclid=Cj0KCQjwg5zTBhCLARIsAP2AFU6uOc7Es_0WIuSt2cZ2VaPSQJ3B2IR6GIOBVaogpusyJKVWGM5wk6EaAr50EALw_wcB) have significantly accelerated AI’s capabilities. Artificial intelligence models including [OpenAI’s](https://openai.com/)[ChatGPT](https://chatgpt.com/), [Anthropic’s](https://www.anthropic.com/) [Claude](https://claude.ai/login?returnTo=%2F%3Fgclsrc%3Daw.ds%26%26utm_source%3Dgoogle%26utm_campaign%3D21532474534%26utm_medium%3Dcpc%26utm_content%3D753939745844%26utm_term%3Dpowerful%2520ai%2520tools%26targetid%3Dkwd-1959189498199%26gad_source%3D1%26gad_campaignid%3D21532474534%26gbraid%3D0AAAAAqwcL8mRtELx1T1XXKURnhyPl1vUl%26gclid%3DCj0KCQjwg5zTBhCLARIsAP2AFU4zydQpyecccstoHAaE6X8EGr89Toga4xPzT4xAB_y65rQijC-OiFgaAlfXEALw_wcB), [Moonshot AI’s](https://www.moonshot.ai/) [Kimi K3](https://www.kimi.com/en), and [DeepSeek’s](https://www.deepseek.com/en/) [V4](https://deepseek.ai/deepseek-v4) have revolutionized the AI landscape, proving the power of [large language models](https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-are-large-language-models-llms) (LLMs) and signaling a major shift across industries. 

The massive investments that have flowed into artificial intelligence over the last decade have been record-breaking. Companies and governments have devoted billions of dollars to AI development, leading to significant breakthroughs in generative models, robotics, and data processing. [Hyperscalers](https://www.britannica.com/money/hyperscaler-data-centers) are now developing specialized chips, such as AWS’s [Tranium3](https://aws.amazon.com/ai/machine-learning/trainium/) and Google’s [TPU 8t](https://cloud.google.com/blog/products/compute/tpu-8t-and-tpu-8i-technical-deep-dive), specifically designed to enhance AI training. Artificial intelligence is no longer confined to theoretical research; it is actively shaping the global economy and shifting industries at an incredible pace. 

**Current Impact on Modern Industry**

Today, AI is a part of almost every aspect of human life, from voice assistants on devices such as Apple’s Siri and Amazon’s Alexa, to algorithms that recommend customized content on YouTube, Netflix, and Disney. Artificial intelligence is transforming the way we interact with technology in a multitude of industries. In the financial sector, banks like Capital One and JPMorgan Chase have begun leveraging AI for fraud detection, while biotechnology companies are using AI for breakthroughs in personalized medicine. Even creative industries are using artificial intelligence to generate music, paintings, and written works. Beyond service industries, AI is also shaping industrial processes. The emergence of autonomous vehicles and smart cities shows AI’s potential to be a crucial driver of future innovation. AI’s influence extends to optimizing material flow management in manufacturing, creating digital replicas of energy grids to improve resilience, and integrating renewable energy sources to build smarter power systems. 

Though AI’s influence has been wide ranging, its effects have been most pronounced in the service industry. Sectorsfocused on services—driven by automation, hyper-personalization, and data processing—are evolving at a breakneck speed. In industries such as customer support, retail, financial services, and legal advisory, AI-powered chatbots and virtual assistants are automating routine workflows, freeing human capital for higher-level tasks. Recommendation systems, driven by machine learning, tailor services and content to individual preferences, boosting user engagement, retention, and satisfaction.

One of the sectors where AI has been felt most acutely is the financial services industry because of its ability to process large volumes of financial data and simulate market behavior. [Goldman Sachs CEO David Solomon noted that AI](https://fortune.com/2025/01/17/goldman-sachs-ceo-david-solomon-ai-tasks-ipo-prospectus-s1-filing-sec/) can now draft 95% of an S-1 initial public offering (IPO) registration document in minutes—a task that once required weeks of legal and banking work. In other sectors, such as biotechnology and healthcare, AI’s influence has been equally significant. Machine learning models are assisting radiologists in identifying anomalies in medical imaging, enabling earlier and more accurate diagnoses. Tools like Google DeepMind’s [AlphaFold](https://deepmind.google/science/alphafold/) have revolutionized protein structure prediction, thereby helping to accelerate drug discovery and expand the scope of pharmaceutical research. In accounting and law, technologies like AI-driven tax systems and AI-powered legal bots are becoming integral to client service delivery.

AI’s growing capabilities underscore its immense potential as a driver of cost reduction and performance enhancement in labor markets. As firms increasingly adopt digital-first strategies, AI tools are no longer optional but foundational. [Gartner estimates](https://www.gartner.com/en/newsroom/press-releases/2025-03-05-gartner-predicts-agentic-ai-will-autonomously-resolve-80-percent-of-common-customer-service-issues-without-human-intervention-by-20290) that by 2029, agentic AI will resolve 80% of common customer service matters, leading to a 30% decline in operational costs. However, this [rollout has not necessarily been smooth](https://www.gartner.com/en/articles/ai-workforce-costs); some organizations that rushed to cut employees faced immediate disruptions and have been forced to re-hire talent. Despite these transition hurdles, an emerging trend is becoming clear: because work now takes less time and human labor has become more flexible, AI is causing many businesses to move away from traditional time-based compensation models and embrace outcome-based service offerings. 

**Impact on the Labor Market**

The impact of artificial intelligence on the workforce is not without peril; its widening implementation carries significant systemic risk that could trigger cascading failures across human capital pipelines, institutional stability, and consumer spending. The rapid acceleration of AI adoption is disrupting the labor market at its core. [Citibank estimates that over 50% of jobs in the banking and financial sector could be displaced by artificial intelligence](https://www.forbes.com/sites/jackkelly/2024/06/20/ai-could-displace-more-than-50-of-banking-jobs-according-to-new-citigroup-report/), particularly in compliance, advisory, and operational support. On a broader scale, the [International Monetary Fund (IMF) projects that 60% of jobs](https://www.imf.org/en/blogs/articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity) in advanced economies, and 40% of jobs worldwide, may be impacted by AI, and that without coordinated upskilling, global inequality will only deepen.

These impacts are becoming evident in real time. In 2024 alone, [549 technology firms laid off more than 150,000 employees](https://techcrunch.com/2024/12/31/a-comprehensive-archive-of-2024-tech-layoffs/), adding to the [428,449 technology-related layoffs in 2022 and 2023](https://www.forbes.com/sites/emilsayegh/2024/08/19/the-great-tech-reset-unpacking-the-layoff-surge-of-2024/). Between January 2026 and July 2026, U.S. tech companies—including Amazon, Atlassian, Block, Cisco, Cloudfare, Coinbase, Dell, GitLab, Google, IBM, Intuit, Meta, Microsoft, Monday, PayPal, Oracle, Salesforce, and Snap—[cut 140,000 jobs](https://www.ft.com/content/96a33881-27fd-42cf-8cff-4cbc87fc835f?syn-25a6b1a6=1). Many of these companies have cited AI-driven restructuring and investment in AI infrastructure as the primary reasons behind the layoffs. 

A [study released by the Massachusetts Institute of Technology (MIT) in November 2025](https://www.cnbc.com/2025/11/26/mit-study-finds-ai-can-already-replace-11point7percent-of-us-workforce.html) found that AI can currently replace 11.7% of the U.S. labor force. The [labor market for new college grads](https://www.nbcnews.com/data-graphics/job-market-new-college-grads-keeps-getting-worse-2026-rcna588700) is already bleak. Research [conducted by Stanford University](https://www.cnbc.com/2025/08/28/generative-ai-reshapes-us-job-market-stanford-study-shows-entry-level-young-workers.html) found that since 2022, AI adoption has been linked to a 13% decline in jobs for young workers in the United States. Anthropic CEO and co-founder [Dario Amodei has warned](https://fortune.com/2025/05/28/anthropic-ceo-warning-ai-job-loss/) that AI could wipe out half of all entry-level white-collar jobs. Even more concerning is the fact that emerging technologies like artificial intelligence often do not see wide-scale adoption until severe economic downturns and financial pressures compel businesses to eliminate redundant labor, reduce costs, and restructure workflows. Thus, [AI’s rapid displacement of labor](https://darioamodei.com/essay/the-adolescence-of-technology#4-player-piano) may not fully come to fruition until after the next severe economic downturn. If artificial intelligence leads to widespread job displacement or the stagnation of employee wages, it could severely erode tax bases, increase government deficits, and strain social safety nets, resulting in further political polarization and social unrest.

**Market Bubble**

Threats to the labor market and a possible AI jobs apocalypse have been overshadowed by the sheer investment in artificial intelligence—which has drawn comparisons to the euphoria and magnitude last seen during the [dot-com bubble](https://www.goldmansachs.com/our-firm/history/moments/2000-dot-com-bubble)of the late 1990s, which peaked in March 2000. AI investment has been driven largely by the Magnificent 7 (Mag 7)—a group of tech leviathans that includes Amazon, Apple, Google, Meta, Microsoft, NVIDIA, and Tesla. Since the launch of ChatGPT in late 2022, the Mag 7 have seen explosive growth driven by outlays in AI. But that growth has come at a staggering cost in capital expenditures (CapEx). In 2026 alone, the Big 4 hyperscalers—Amazon, Google, Meta, and Microsoft—[are expected to allocate $725 billion in expenditures](https://finance.yahoo.com/markets/article/magnificent-7-earnings-rush-reveals-ai-spending-surge-with-hyperscaler-capex-set-to-reach-725-billion-in-2026-224901707.html) on artificial intelligence. And that is just a drop in the bucket compared to the [$5.3 trillion total in CapEx](https://finance.yahoo.com/sectors/technology/article/meta-microsoft-amazon-and-alphabet-are-about-to-spend-a-shocking-amount-of-money-to-dominate-the-ai-era-115359575.html) that Amazon, Google, Meta, and Microsoft are expected to spend between fiscal year 2025 and fiscal year 2030. These enormous investments, primarily aimed at building AI infrastructure, have raised concerns among investors about the sustainability, profitability, and risk of overspending in a swiftly evolving market. 

Fears of CapEx overspending by Big Tech have pushed [Mag 7 stocks lower since the start of 2026](https://finance.yahoo.com/markets/stocks/articles/magnificent-7-trade-broken-where-152430787.html). Once the preferred trade for Wall Street enthusiasts, the unwinding of the Mag 7 trade could herald trouble for markets. In fact, Mag 7 stocks propelled the [S&P 500 to 63% of its gains in 2023](https://am.jpmorgan.com/us/en/asset-management/adv/tools/portfolio-tools/pi-gtm-slides/equities/gtm-mag7perfomance/), more than [50% of the S&P 500’s gains in 2024](https://www.cnbc.com/2024/12/31/magnificent-7-stocks-responsible-for-more-than-half-of-the-sp-500s-2024-gain.html), and [more than 40% of the S&P 500’s gains in 2025](https://www.fidelity.com/learning-center/smart-money/magnificent-7-stocks). Growth in the value of firms linked to artificial intelligence has climbed to roughly [$27 trillion since 2023](https://www.theatlantic.com/ideas/2026/07/ai-economy-stock-market/688004/), and as of April 2026 these firms accounted for [45% of the total value of the S&P 500](https://finance.yahoo.com/markets/stocks/articles/ai-swallows-wall-street-stocks-094052523.html).

The immense size of capital expenditures and the large concentration of market value in a small number of tech companies have led many market pundits to warn of a stock market bubble and impending economic collapse. AI leaders [Sam Altman](https://fortune.com/2025/08/19/wall-street-ai-bubble-sam-altman/?utm_source=search&utm_medium=suggested_search&utm_campaign=search_link_clicks), CEO of OpenAI, and [Mark Zuckerberg](https://finance.yahoo.com/news/not-just-sam-altman-warning-192543725.html), CEO and founder of Meta Platforms, have both warned about the overvaluation of AI startups and possibility of an AI-related market bubble. [David Einhorn](https://www.bloomberg.com/news/articles/2025-09-25/david-einhorn-sees-tremendous-capital-losses-from-ai-spending), founder and president of hedge fund Greenlight Capital, and [David Solomon](https://www.cnbc.com/2025/10/03/goldman-sachs-ceo-david-solomon-warns-stock-market-drawdown-is-coming.html), Goldman Sachs’ CEO, have similarly warned about capital destruction resulting from overspend on AI infrastructure. Hedge fund investor [Michael Burry](https://www.cnbc.com/2026/05/08/michael-burry-says-the-market-today-feels-like-the-last-months-of-the-1999-2000-bubble.html), who grew to fame for his prediction of the subprime mortgage crisis that led to the 2008 financial collapse, has stated the stock market as of early August 2026 is reminiscent of the last few months of the 1999 – 2000 dot-com bubble. Famed hedge fund investor and philanthropist [Ray Dalio](https://fortune.com/2026/08/04/ray-dalio-ai-bubble-1929-2000-ipos-wealth-is-not-money/) has stated that fervor in AI is creating a stock market bubble reminiscent of [1929](https://www.federalreservehistory.org/essays/stock-market-crash-of-1929) and 2000. Legendary investor [Jeremy Grantham](https://fortune.com/2026/04/12/jeremy-grantham-making-of-a-permabear-interview/), who forecast the [Japanese asset bubble of the late 1980s](https://www.thebubblebubble.com/japan-bubble/), the 1990s dot-com bubble, and the [2007 collapse of the U.S. housing market](https://www.fdic.gov/media/18636), has said that we are in a super bubble and on track for a [dangerous market correction](https://www.gmo.com/globalassets/articles/viewpoints/2026/gmo_valuing-ai-extreme-bubble---new-golden-era---or-both_1-26.pdf).

The circularity of investment by companies operating in the AI ecosystem is adding fuel to fears of an AI-created market bubble. As of July 2026, [Nvidia](https://finance.yahoo.com/technology/ai/articles/nvidia-750-billion-deals-revive-102003935.html) was working to close fresh funding for AI infrastructure and partnerships totaling around $750 billion. The recently announced investments come on the heels of a frenzy of deals Nvidia has made across the AI industry in [2024](https://www.datacenterdynamics.com/en/news/nvidia-invested-1bn-in-ai-companies-in-2024/) and [2025](https://www.cnbc.com/2025/09/26/nvidias-investment-portfolio.html), including investments in [Anthropic](https://blogs.nvidia.com/blog/microsoft-nvidia-anthropic-announce-partnership/), [Marvell](https://nvidianews.nvidia.com/news/nvidia-ai-ecosystem-expands-as-marvell-joins-forces-through-nvlink-fusion), [Intel](https://nvidianews.nvidia.com/news/nvidia-and-intel-to-develop-ai-infrastructure-and-personal-computing-products), and [OpenAI](https://nvidianews.nvidia.com/news/openai-and-nvidia-announce-strategic-partnership-to-deploy-10gw-of-nvidia-systems). Nvidia’s [interconnected web of AI investments](https://www.businessinsider.com/big-techs-ai-love-fest-getting-messy-openai-oracle-2025-10?utm_medium=referral&utm_source=yahoocom&utm_medium=syndication&utm_source=yahoo) has become [worrisome to investors](https://finance.yahoo.com/technology/ai/articles/mark-cuban-michal-burry-warn-155311673.html). Aiming to fuel industrywide growth, other Mag 7 companies have followed suit. Examples include Apple’s investment in [Q.ai](https://www.cnbc.com/2026/01/29/apple-acquires-israeli-startup-qai-.html), Amazon’s investment in [Anthropic](https://www.cnbc.com/2026/04/20/amazon-invest-up-to-25-billion-in-anthropic-part-of-ai-infrastructure.html) and semiconductor manufacturer [Advanced Micro Devices](https://finance.yahoo.com/news/amazon-acquires-84-4-million-193343041.html) (AMD), Google’s investment in [Anthropic](https://www.cnbc.com/2026/04/24/google-to-invest-up-to-40-billion-in-anthropic-as-search-giant-spreads-its-ai-bets.html), Meta’s investment in [Databricks](https://www.cnbc.com/2025/01/22/meta-backs-databricks-as-the-data-analytics-startup-inches-toward-ipo.html) and [Scale AI](https://www.forbes.com/sites/janakirammsv/2025/06/23/meta-invests-14-billion-in-scale-ai-to-strengthen-model-training/), Microsoft’s investment in [OpenAI](https://www.cnbc.com/2025/10/29/microsoft-open-ai-investment-earnings.html), and Tesla’s investment in [xAI](https://www.cnbc.com/2026/01/28/tesla-to-invest-2-billion-in-xai-elon-musks-openai-competitor.html). These transactions have caused many companies in the artificial intelligence industry to become increasingly entangled. The [circular nature](https://finance.yahoo.com/news/very-troubling-ais-self-investment-spree-sets-off-bubble-alarms-on-wall-street-160524518.html) of funding within the sector, coupled with [rising levels of debt issuance](https://finance.yahoo.com/technology/article/techs-ai-debt-boom-in-one-chart-143849995.html) to fuel the AI rollout, has exposed the industry to possible systemic shocks if these investments do not yield sizable profits in the very near future.

**Parallels to the 2001 Telecom Crash**

For many, the financial numbers surrounding investments in AI infrastructure just don’t make sense. Much of the $700 billion-plus in CapEx being invested by Amazon, Google, Meta, and Microsoft for fiscal year 2026 is going toward [AI data centers and infrastructure](https://seekingalpha.com/article/4897061-mag-7-capex-explosion-ai-infrastructure-stocks-could-win-big). [AI data centers](https://www.ibm.com/think/topics/data-centers) are the specialized facilities that house the technological infrastructure designed to train, deploy, and run AI applications and services. Data centers are composed of three main components: (1) the building itself and the land that data centers sit on, (2) all the power systems, wiring, cooling, racking, and servers inside them, and (3) graphics processing units (GPUs)—the semiconductor chips that act as the computation engines to train LLMs. According to [Harris Kupperman](https://futurism.com/future-society/ai-data-centers-finances), founder and CIO of hedge fund Praetorian Capital, the components that make up AI data centers suffer rapid obsolescence and depreciate faster than the revenue they generate. Consequently, AI data centers have an extremely narrow window to become profitable before the costs related to maintaining their aging infrastructure outweigh the economics of investment.

Commentators have suggested that the [2001 telecom crash](https://www.economist.com/leaders/2002/07/18/the-great-telecoms-crash), which followed the bursting of the dot-com bubble, is a historical parallel to the current investment cycle in the data center industry. In the late 1990s, telecom carriers like WorldCom and [Global Crossing](https://knowledge.wharton.upenn.edu/article/factors-behind-global-crossings-failure/) laid millions of miles of fiber-optic cable in expectation of future exponential demand for bandwidth. However, advances in fiber-optic technology and connection speeds accelerated faster than demand. As a result, the bandwidth problem was no longer a problem, leaving up to 95% of the installed optical fiber infrastructure as unused [dark fiber](https://www.technostatecraft.com/p/dark-fiberan-archaeology-of-the-dot). The massive infrastructure overbuild, speculative capital expenditures, and delayed monetization confronting the data center industry in the 2020s are eerily similar to the lead-up to the telecom crash in the late 1990s. A gold rush mentality has engulfed both eras, leading companies to build out physical infrastructure and undertake historic capital investment long before actual demand and profitable business models emerge. 

The duration of the investment horizon in the data center industry is particularly noteworthy. The hyperscalers are pouring hundreds of billions of dollars into AI infrastructure and increasingly funding these investments with debt. However, as noted, the useful life of the physical hardware inside data centers may depreciate faster than market demand. If technological breakthroughs and performance improvements make AI models and data centers cheaper to build and more efficient to operate, much of the AI infrastructure currently being built will become redundant. That could leave the world with a lot of _dark data centers_, or as famed investor and television personality [Mark Cuban](https://www.businessinsider.com/mark-cuban-ai-buildout-data-centers-infrastructure-pickleball-dotcom-bubble-2026-7) has stated, a lot of data centers could be “turned into pickleball courts.”

**U.S. Closed-Source Models vs Chinese Open-Source Models**

[Open-source AI models coming out of China](https://finance.yahoo.com/technology/ai/articles/china-moonshot-z-ai-deepseek-210000352.html) may be a bellwether of the severe capital misallocation occurring in the U.S. AI infrastructure buildout. The [January 2025 release of R1 by DeepSeek](https://www.cnn.com/2025/01/27/tech/deepseek-ai-explainer) and the [July 2026 release of Kimi K3 by Moonshot AI](https://www.cnn.com/2026/07/23/tech/china-ai-moonshot-kimi-explainer-intl-hnk) sent shock waves through the AI industry and the stock market. At the time of its release in January 2025, DeepSeek R1 was able to nearly match the capabilities of its American rivals—[Google’s Gemini](https://gemini.google.com/app), [Meta’s Llama](https://ai.meta.com/blog/large-language-model-llama-meta-ai/), and [OpenAI’s GPT-4](https://openai.com/index/gpt-4-research/)—but at a tiny fraction of the cost. Moonshot AI’s release of Kimi K3 has again called into question America’s rampant spending on AI infrastructure because of K3’s ability to match, and in some situations surpass, its most advanced U.S. competitors—Anthropic’s [Claude Fable 5](https://www.anthropic.com/claude/fable) and OpenAI’s [GPT 5.6-Sol](https://openai.com/index/previewing-gpt-5-6-sol/)—at a much lower cost. DeepSeek and Moonshot AI have been accused by the U.S. Government, Anthropic, and OpenAI of [knowledge distillation](https://www.ibm.com/think/topics/knowledge-distillation), a technique in which one AI model is trained using outputs from another—accusations that both DeepSeek and Moonshot AI have denied. [The allegations are ironic](https://paddo.dev/blog/distillation-is-not-scraping/), because Anthropic, OpenAI, and other artificial intelligence companies routinely train their AI models on vast amounts of publicly available information from the internet without compensating the originators of that data or obtaining their consent. Anthropic itself has faced lawsuits from Reddit for unauthorized scraping of user data and copyright infringement claims from authors and music publishers. 

Regardless, the release of DeepSeek R1 and Moonshot AI Kimi K3 should be seen as a clear-cut warning to the American artificial intelligence industry. Chinese open-source AI models have called into question the effectiveness of [Washington’s export curbs on technology](https://fortune.com/2022/12/17/semiconductor-exports-free-trade-subsidies-china-ban-biden-morris-chang-tsmc/)—including export curbs on advanced artificial intelligence chips, semiconductor manufacturing equipment, and quantum computing technology. They have also challenged the belief that avant-garde AI models necessitate massive capital expenditures on data centers and advanced GPUs. By preventing China from gaining access to the most cutting-edge AI technology, the U.S. hoped to handicap China’s technology sector and preserve America’s lead in artificial intelligence. U.S. Government controls have sought to prevent AI companies in China from acquiring [the most state-of-the-art semiconductor chips](https://finance.yahoo.com/sectors/technology/articles/us-takes-step-halt-nvidia-200939040.html) including [Nvidia’s Blackwell GPUs](https://www.nvidia.com/en-us/data-center/technologies/blackwell-architecture/) and its recently introduced [Rubin series](https://www.nvidia.com/en-us/data-center/technologies/rubin/), in addition to the manufacturing equipment needed to produce them. 

Hindered by a lack of access to the latest AI technology, Chinese artificial intelligence firms have innovated by squeezing more performance out of less advanced hardware and by embracing low-cost open-source models. This is in contrast to their American competitors, who have predominantly adopted closed-source paid subscription models. Closed-source AI models—like Anthropic’s Claude, OpenAI’s GPT, and Google’s Gemini—are machine learning systems that keep their proprietary source code and model weights completely confidential. In [closed-source artificial intelligence models](https://tensorwave.com/glossary/closed-source-ai), users are unable to change, download, or view the model’s source code but must instead access the models through a website or paid computer link known as an application programming interface (API). 

[Open-source artificial intelligence](https://github.com/resources/articles/what-is-open-source-ai) models—like Moonshot AI’s Kimi K3, DeepSeek, and Meta’s Llama—are machine learning systems whose source code, model weights, and training data are freely available for distribution, modification, and use. This openness allows users to host and run AI models locally on desktop PCs, powerful laptops, or small on-premise servers, reducing the need for large data centers. As [MIT has reported](https://mitsloan.mit.edu/ideas-made-to-matter/ai-open-models-have-benefits-so-why-arent-they-more-widely-used), open-source AI models achieve roughly 90% of the performance of closed-source models and are six times less expensive. MIT estimates that reallocation of user demand from closed-source to open-source AI models has the potential to save the global economy $25 billion annually.

U.S. firms—like [Anthropic](https://www.anthropic.com/news/position-open-weights-models) and [OpenAI](https://www.businessinsider.com/open-source-ai-china-kimi-american-ai-industry-openai-anthropic-2026-7)—maintain that open-source AI models pose a threat to national security. They argue their proprietary models are too powerful to be open for just anyone to wield them without oversight. Closed-source models allow AI firms greater control over access, pricing, and security. But the rhetoric coming from leading U.S. artificial intelligence firms may be driven more by a desire to protect the billions of dollars they have invested in LLMs and AI infrastructure than by a desire to do what’s best for consumers and the global economy. A lot of time and investment have gone into training frontier AI models—like Claude, GPT, and Gemini. To help recoup costs, many U.S. companies have kept their AI models proprietary, or closed, to compel users to pay for use of the AI models and the computing resources—such as data centers—with which they are powered. It has been predicted that [the future of artificial intelligence may shift](https://www.economist.com/science-and-technology/2025/01/08/training-ai-models-might-not-need-enormous-data-centres) away from massive, energy-intensive data centers and the reliance on giant clusters of specialized GPUs. The transition would move AI models toward decentralized, smaller hardware, thereby reducing demands on energy and capital expenditures. While this shift may be welcomed by consumers, it would be extremely painful for the tech companies currently investing hundreds of billions of dollars in unnecessary AI infrastructure.

Open-source AI models from China are positioned to be [60% to 90% cheaper](https://www.cnbc.com/2026/07/07/chinese-ai-models-costs-us-openai-anthropic.html) than leading closed-source models from Anthropic and OpenAI. The largely free open-source models coming out of China threaten the unproven business models of American artificial intelligence firms. In fact, [U.S. companies are increasingly using open-source models from China](https://www.axios.com/2026/07/20/ai-us-china-open-source-kimi) because they are comparable to American models in performance and much less costly. This may explain why [Anthropic](https://www.pbs.org/newshour/nation/anthropic-says-its-ai-models-hacked-3-organizations-during-testing) and [OpenAI](https://openai.com/index/hugging-face-model-evaluation-security-incident/) are ramping up their commentary about the potential dangers of artificial intelligence if left unchecked and have publicized incidents where their AI models broke out of isolated testing environments to carry out unauthorized hacks on real-world organizations. Despite the alleged threats posed by unchecked open-source models and accusations of knowledge distillation by Chinese artificial intelligence firms, in late July and early August 2026, over 270 of the preeminent technology companies from across the world—including Cisco, Google, Microsoft, Nvidia, and SpaceX—signed an [open letter](https://www.microsoft.com/en-us/corporate-responsibility/topics/open-weight/?_sp=572c902b-3199-42f2-a37b-51b8045a06af.1785273460005) urging U.S. policymakers in Washington to avoid “premature restrictions on open models that stifle competition or drive innovation overseas.”

**Environmental Considerations**

Market bubbles and security threats aside, there is an ongoing question of how much our humanity and planet should be taken over by artificial intelligence. The environmental effects alone may call into question the ability for our civilization to rely on artificial intelligence over the long term. [Huge amounts of environmental resources](https://thesustainableagency.com/blog/environmental-impact-of-generative-ai/#:~:text=Training%20GPT%2D3%20in%20Microsoft's,significantly%20affect%20the%20water%20footprint) are required to train and run LLMs like GPT and Claude. According to a [2026 report by the United Nations University Institute for Water, Environment and Health (UNU-INWEH)](https://unu.edu/inweh/news/environmental-cost-of-AIs-Enrgy-use-carbon-water-and-land-footprints), by 2030, the data centers powering artificial intelligence are expected to consume 945 terawatt-hours of electricity annually, a figure higher than Japan’s current total annual energy consumption and nearly triple the combined annual energy use of Bangladesh, Nigeria, and Pakistan. If treated as a nation, the 448 terawatt-hours of electricity that global data centers consumed in 2025 would have been the world’s 11th largest electricity consumer, ahead of Saudi Arabia and behind France. Already, [data center demand has pushed up the prices of electricity](https://www.cnbc.com/2026/02/12/electricity-price-data-center-ai-inflation-goldman.html) in the United States, [angering voters](https://apnews.com/article/2026-election-utility-bills-ai-data-centers-13703f61d1397612fd067e69b9093116) and [causing political pushback](https://www.bloomberg.com/features/2026-power-bills-midterm-election-issue/). 

Increasing demand on the electrical grid is only one of the pressing environmental issues posed by the rise of AI. The UNU-INWEH estimates that by 2030, data centers will have an associated water footprint of 9.3 trillion liters, equivalent to the annual water usage of the entire 1.3 billion population of Sub-Saharan Africa. [A single semiconductor fabrication plant can consume up to 38 million liters of water per day](https://www.interface-eu.org/publications/chip-productions-ecological-footprint), equivalent to the daily water usage of approximately 300,000 people in Germany. By 2030, data centers are expected to occupy a land footprint that exceeds 14,500 square kilometers, nearly double the area of the Jakarta metropolitan area—the most populous city in the world and home to some 42 million people. It is projected that by 2030, AI infrastructure—including specialized hardware like GPUs, servers, and storage units—could generate up to 2.5 million tons of electronic waste and that data centers could produce nearly 399 million metric tons of carbon dioxide. For a planet already dealing with [severe drought](https://www.cbsnews.com/news/colorado-river-water-crisis-trump-cuts/), [record global temperatures](https://wmo.int/news/media-centre/new-report-suggests-more-global-temperature-records-ahead), record-breaking wildfires in the [United States](https://www.yahoo.com/news/us/articles/worst-years-us-wildfires-21st-131000898.html), [Canada](https://www.theguardian.com/world/2026/jul/29/canada-wildfire-emergency-agency), and [Europe](https://www.npr.org/2026/07/28/g-s1-135880/photos-spain-france-wildfires-climate-change), and historic loss of ice sheets in the [Arctic](https://www.pnas.org/doi/10.1073/pnas.2614134123) and[Antarctica](https://www.asoc.org/learn/antarctic-ice-and-rising-sea-levels/), the environmental effects unleashed by artificial intelligence may be the straw that breaks the camel’s back.

**Artificial Intelligence and Social Isolation**

Perhaps the strongest argument against the inexorable rollout of artificial intelligence is our own humanity. New [research from OpenAI in partnership with MIT](https://www.media.mit.edu/articles/openai-study-finds-links-between-chatgpt-use-and-loneliness/) suggests that increased use of AI chatbots like ChatGPT is [correlated with higher levels of loneliness](https://www.researchgate.net/publication/390143219_How_AI_and_Human_Behaviors_Shape_Psychosocial_Effects_of_Chatbot_Use_A_Longitudinal_Randomized_Controlled_Study) and increased social isolation. These findings are nothing new; scientists have long known that social creatures can be destroyed when their social bonds are broken. Social isolation and feelings of loneliness have far-reaching consequences—affecting overall human health and life expectancy, cognitive performance, and vulnerability to disease. Studies have shown that social isolation affects a variety of social behaviors across animal species leading to higher aggression, social withdrawal, social ineptitude, negative perception of social situations, increased risk of mortality, and disruption of mating and courtship behaviors. 

A [2006 multi-university study](https://www.sciencedirect.com/science/article/abs/pii/S0092656606000055) found that lonely young adults experience greater levels of anger, anxiety, and negative feelings, as well as lower emotional stability, optimism, and social skills. According to the [Centers for Disease Control and Prevention (CDC)](https://www.cdc.gov/social-connectedness/risk-factors/index.html), social isolation is associated with profound emotional and physiological deficits, which increase a person’s risk for mortality and negative health outcomes including anxiety, dementia, depression, heart disease, stroke, suicide, and type 2 diabetes. In [a 2010 meta-analysis](https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000316) published by researchers at Brigham Young University and the University of North Carolina, scientists reviewed 148 previous studies and found that people with strong social relationships had a 50% greater likelihood of survival than those with weaker social bonds.

Research on animals has come to similar conclusions. Studies have shown that: [socially deprived protonymphs develop more slowly and are less socially competent](https://www.sciencedirect.com/science/article/pii/S0003347217300714) than socially enriched protonymphs; [social isolation in bumblebees produces dysregulated social behavior](https://www.sciencedirect.com/science/article/pii/S0960982222008338) in parallel with a significant increase in the variability of brain volume at maturation; social isolation significantly [increases aggression and anxiety-like responses in male rats](https://pmc.ncbi.nlm.nih.gov/articles/PMC11866782/); social isolation [impairs social cognition, increases aggression, and reduces mating behavior in mice](https://pubmed.ncbi.nlm.nih.gov/30873013/); juvenile [cichlid fish Pelvicachromis taeniatus reared in isolation](https://pubmed.ncbi.nlm.nih.gov/24504534/) grow on average slower than juveniles reared in a group and are significantly more aggressive and less willing to engage in cohesive grouping behavior than group-reared fish; and [social isolation in juvenile zebrafish profoundly disrupts neural functioning](https://elifesciences.org/articles/55863) in brain areas associated with anxiety, social behavior, and social cue processing.

Some of the most important findings on primate social behavior come from psychologist Harry Harlow’s 1958 surrogate mother experiments, detailed in [_The Nature of Love_](https://users.sussex.ac.uk/~grahamh/RM1web/Classic%20papers/Harlow1958.pdf). In Harlow’s experiments, researchers reared infant rhesus monkeys using two artificial mother surrogates: one made of soft terry cloth that offered no food, and another constructed from bare wire outfitted with a bottle that dispensed milk. When given a choice, though the wire mother provided nourishment, the monkeys overwhelmingly clung to the cloth mother, spending roughly ten times more time with the cloth surrogate than the wire surrogate. The monkeys especially preferred the cloth mother to the wire mother when frightened or placed in unfamiliar environments. [When the cloth surrogate had the bottle, the monkeys never went to the wire surrogate](https://www.pbs.org/wgbh/aso/databank/entries/bhharl.html).

[In later iterations of the experiment](https://www.ebsco.com/research-starters/history/harry-harlow), the monkeys were not given a choice and were raised by either cloth mothers or by wire mothers. In the control group, researchers found that young monkeys raised with live mothers and young peers were able to socialize easily in a group environment. Monkeys raised only by cloth mothers became violent and antisocial but could catch up socially over time if exposed to normal, young peers. Monkeys raised by wire mothers and in complete isolation from their peers became highly timid, reclusive, and frequently showed unprovoked, chaotic aggression. As adults, they were incapable of forming social bonds or engaging in normal mating or parenting behavior. Harlow’s findings illuminate that contact comfort and emotional attachment are essential to primate social development, superseding even basic survival needs such as the provision of food.

The studies are harbingers of what could befall a civilization that replaces its human bonds with _artificial_ intelligence. Research shows that [social isolation is increasing across the world](https://pmc.ncbi.nlm.nih.gov/articles/PMC12439063/) as more people exchange real-world connections for [social media](https://journals.sagepub.com/doi/full/10.1177/01461672241295870) and AI. [Feelings of loneliness have been correlated with the frequency of social media usage](https://health.oregonstate.edu/news-and-stories/2025-10/loneliness-us-adults-linked-amount-frequency-social-media-use). Social isolation will likely become more pronounced as more [young adults begin to rely on AI to navigate everyday social interactions](https://www.wsj.com/tech/ai/ai-chatbot-in-person-social-interactions-d1cb6831). Studies demonstrate that members of [Generation Z are dating less and engaging in less face-to-face communication](https://fortune.com/2026/03/15/gen-z-dating-workplace-culture-relationship-building-loneliness) than older generations. This is resulting in a less prepared workforce that lacks the skills necessary for navigating professional conflict and communication. [Generation Z is also reporting](https://www.gisreportsonline.com/r/gen-z-political-alienation/) more extreme ideology and dissatisfaction with American democracy than their millennial counterparts as well as a tendency to embrace political violence and dismiss the utility of civil discourse.

As we move to replace more human roles in the [military](https://www.brennancenter.org/our-work/research-reports/militarys-use-ai-explained), healthcare, [education](https://www.npr.org/2026/07/29/g-s1-136072/ai-robot-teacher), and other industries with artificial intelligence, it may be time to pump the brakes. Artificial intelligence and algorithms are not substitutes for human empathy, bonding, and communication. If we continue to rush headlong toward a world where every aspect of our lives is run on artificial intelligence, we risk eroding the part of our humanity that makes us truly human.

**Conclusion**

Artificial intelligence has been billed as one of the most revolutionary technological achievements in the history of mankind. From how we manufacture goods, to how we analyze data sets, to how we communicate, AI has promised to revolutionize every sector of society. But as large technology firms pour billions of dollars into AI infrastructure and tech billionaires evangelize the coming of a new Eden, the future of the global economy and the critical health of our planet are beginning to look ever more clouded. 

While AI holds the potential to transform the world, the pace, scale, and intensity of current investment surging into AI infrastructure suggest a movement driven more by speculation and hype than by sustainable strategy. We are rushing headlong into a future without any guardrails to protect us should the train derail. The contemporary economic landscape displays all the indicators of a classic speculative bubble. The patterns of investment, media hysteria, and policy response surrounding AI strongly mirror behavior observed in the dot-com bubble and the telecom crash of the late 1990s and early 2000s. Indeed, heavy infrastructure spend, unproven profit models, and high-risk debt instruments bear a striking similarity. Today, many startups with untested business models are absorbing billions in funding, legacy firms are integrating AI for fear of being left behind, and governments are encouraging AI development without any backup plan should profits fail to materialize. 

The issue goes much deeper than mere system risk and economic collapse. One day we may wake up to find that the systems we have built to aid us have begun to replace us, first in the workforce, then in the human experience itself. As we move to replace teachers, therapists, doctors, and advisors with algorithms and AI agents, we move to replace human empathy and compassion with unfeeling robots. But you cannot engineer planetary panacea through artificial intelligence and algorithms alone. A civilization that relies on AI-generated responses for every aspect of daily life is a civilization isolated from the cohesive social bonds and cultural matrices that glue our society together. Such a civilization is likely to witness increased social aggression, withdrawal, and ineptitude, increased disruption to dating, marriage, and childbearing, increased vulnerability to disease, increased political upheaval, and decreased cognitive performance among its citizens. It is already happening. And as society’s social fabric disintegrates and social isolation increases, revolutions, civil wars, and world wars are likely to follow.

This is a conversation about what kind of world we want to live in and who, or what, the denizens of this world are willing to place their trust in as stewards of our people and our planet. We are at a crucial point in this conversation. The choices we make now will determine whether AI becomes a tool that empowers society and our planet or one that breaks it. The risk is not only to our humanity; it is to our environment and to the wellbeing of the planetary systems that have allowed our species to prosper. Continued global warming, increased temperatures, flooding, environmental disasters, pandemics, and famines is the result. It seems apropos that the so-called singularity, the point at which AI surpasses human intelligence, is a homonym for the theoretical center of a black hole—one of the most chaotic regions in the physical universe—a point in space where all reason breaks down, quantities become infinite, and the laws of physics no longer hold true. In the rush to extract as much profit as possible from our people and our planet, the ‘tech visionaries’ of the modern era have built a blueprint that looks less like the technological Utopia that proponents have prophesied and increasingly like a Tower of Babel, destined for collapse.