"Happy 2025. I lost my freelance writing job to AI,” YouTuber Alex Wei titled a video that went viral on New Year's Day. In the video, he details how a client dumped him in favor of using an AI chatbot to crank out blog posts.

“How can I compete with that?” he asked.

For Wei and millions like him, the path to staying competitive is not at all clear—even for people who know how to use AI to avoid being replaced by AI. And for those who do manage to use AI to stay ahead of the career wrecking ball, it’s getting increasingly costly and difficult to hold onto jobs, especially in the developing world.

OpenAI's latest "pro" tier subscription costs $200 per month. RunwayML (a leading video generator) charges $95 monthly for its premium features, while the best Midjourney (an AI image generator) tier runs at $120 per month. Just a year ago, OpenAI’s top tier for ChatGPT Plus was priced at around $20, with Runway charging $15 to grant access to its Gen-3 video generator.

While $200 might seem reasonable for a business tool in the U.S, it represents around two months of an average minimum wage in Venezuela, equals two weeks' pay in Mexico or Chile, and matches the monthly minimum in Suriname.

Even in emerging economies like China, where the monthly minimum wage varies by region from $275 to $370, these subscription costs can consume a significant portion of a worker's income—especially if they are freelancing.

The AI haves and have-nots

These wallet-busting prices are creating schisms between those who can afford to harness AI's power and those left watching from the sidelines. Daniel Vasilevski, who runs an electric company in Australia called Bright Force Electrical and pays $120 monthly to use Midjourney for his business, sees the writing on the wall.

"The impact that I see here is that AI would widen the gap between companies that can purchase it and those that cannot," Vasilevski told Decrypt. "Companies that purchase advanced AI would perform better in automating work, assisting their clients, and making decisions, whereas small companies or individuals that cannot purchase it would struggle to compete."

Added Vasilevski: “If access is based on budget, it will concentrate all the power in the hands of those who can afford it, leaving others at a disadvantage.”

Jeff Le, former deputy cabinet secretary for California who oversaw emerging tech portfolios for Governor Jerry Brown, sees some parallels between those times and the current status quo, but is still cautious about the future.

"The tools could change the way we all do work and create opportunities for more innovation. But it still seems premature and still in the hands of the few," Le told Decrypt.

New technology concentrating wealth and power in the hands of a few is hardly a new story. The Gini index measures how the gap between the rich and the poor in a country grows over time. With the advent of the Internet, even though GDP grew across the board, the Gini index went up, showing that the gap in opportunities and wealth distribution between rich and poor nations widened.

In other words, technology made countries richer, but did not necessarily make their poor populations less poor. The GDP grew collectively thanks to the globalization of the markets and the adoption of new technologies, but in reality, the profit went to a smaller amount of people—only increasing the gap between the wealthy and the poor.

Can regulations brook the divide?

The situation mirrors what happened after the Telecommunications Act of 1996 in the U.S., when market-driven solutions prioritized urban and affluent areas over rural and low-income communities. By 1999, only 9% of U.S. classrooms had internet access—typically in the richest school districts—leading civil rights leader Jesse Jackson to condemn what he called technological segregation.

The U.S. Congress is paying attention. A recently established bipartisan House AI Task Force examined how to prevent AI from widening societal gaps, much like lawmakers did with internet access in the '90s. But just like the internet's early days, when the price for an AOL subscription seemed high, today's AI tools command premium prices that can become prohibitively steep as AI is more widely adopted.

The result may be a deepening innovation gap. For example, AI-driven healthcare diagnostics are widely deployed in the U.S. but remain rare in low-resource settings, due to high compute costs and data scarcity. Furthermore, regulatory hurdles—such as the EU’s AI Act—disproportionately burden smaller players, stifling local innovation.

The problem could sort itself out over time, of course. Among academic researchers, there seems to be consensus that though the burden to invest in AI adoption is inevitably higher among poorer countries, it is beneficial in the long run.

“While technological catch-up is attainable, it necessitates meticulous planning, investments in human capital, and policy interventions,” according to a recent study in Nature. “The absence of requisite digital infrastructure, skilled workforce, and research capabilities often hinders direct AI advancement pathways for LICs (low income countries).”

However, “evidence shows that technologies like mobile-based e-commerce and e-banking have been adopted faster in low- and middle-income countries (LMICs) compared to HICs, supporting the idea that some LICs can leapfrog in AI adoption with the right conditions.”

Regulators may not have the last word

Without targeted interventions, such as subsidized access to open-source models or hybrid cloud solutions, AI risks becoming another axis of global inequality, mirroring the early internet’s exclusionary dynamics.

And some believe this is a systemic issue that can’t be tackled with regulations alone—the market itself will find a solution.

Increased competition could ultimately drive prices down. And open-source solutions, such as China’s DeepSeek R1, which utterly humiliated OpenAI, could also level the playing field. Beyond its open source model, DeepSeek offers power users a language model at just $0.07 per million tokens—a fraction of GPT-4's $2.50 price tag. The company fired a shot across the bow of industry giants, demonstrating that high prices stem more from market monopolization than actual computing or efficient R&D costs.

Consequently, OpenAI released its beefy reasoning model for the cheaper “Plus” tier, Perplexity adopted a local version of R1 for western users and released a deep research model, and reports emerged that Anthropic was also working on a reasoning model to stay competitive.

“Market forces will address AI accessibility more effectively than corporate mandates,” Karan Sirdesai, CEO and co-founder of AI infrastructure company Mira Network, told Decrypt. “More companies are building open-source alternatives to premium AI tools, creating competition that benefits SMEs. This natural evolution toward accessible solutions mirrors how other technologies have become democratized through market dynamics rather than regulation.”

Even OpenAI CEO Sam Altman is trying to think outside the box with solutions that involve promoting AI among the underserved:

“In particular, it does seem like the balance of power between capital and labor could easily get messed up, and this may require early intervention” he wrote on his official blog. “We are open to strange-sounding ideas like giving some ‘compute budget' to enable everyone on Earth to use a lot of AI.”

This, of course, is still far from ideal as it would only increase users’ dependency on OpenAI tools, further strengthening the company’s monopoly. Whether open-source alternatives, regulatory action, or sheer market competition can balance the scales remains to be seen—but for now, the AI revolution is anything but evenly distributed.

“At its core, regulation must strike a balance between mitigating risks and fostering innovation, ensuring AI does not become a resource exclusive to the wealthy and powerful,” Atul Arya, CEO and founder of AI software provider Blackstraw.ai, told Decrypt.

“We must prioritize equitable access to the infrastructure, talent, and funding necessary to develop AI solutions," he added. "Open innovation ecosystems, public-private partnerships, and initiatives to lower the barrier of entry for custom AI development will play a critical role in ensuring that the benefits of AI are broadly shared.”

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。