In 1879, economist Henry George posed this question in Progress and Poverty, observing that industrial advances enriched some but left many behind. His era’s culprit? Land monopoly – the way rising land values let a few siphon off the gains of growth.
George’s solution was radical for its time: a single tax on land to reclaim that unearned wealth for society. Today, as we stand in the Intelligence Age, a similar paradox looms.
AI-driven progress is causing many to ask whether this new wealth will be broadly shared or concentrated in a few hands. Will automation uplift everyone, or will we once again see great progress alongside great poverty?
Progress, but for Whom?
Today, AI is automating tasks, generating content, and altering decision-making in boardrooms across the Fortune 500.
Tech optimists project a significant boost to global growth: Goldman Sachs predicts generative AI will lift world GDP by 7% over the next decade, and McKinsey’s analysts find [AI could raise annual GDP growth rates by 3–4 percentage points through 2040](https://www.mckinsey.com/capabilities/mckinsey-digital/our-insights/the-economic-potential-of-generative-ai-the-next-productivity-frontier#:~:text=Generative AI’s impact on productivity,entire GDP in 2021 was).
In short, we may be on the cusp of a new productivity bonanza – a “blue-collar bonanza,” as The Economist has called it – driven by intelligent automation. Today, we also see signs that AI’s benefits might pool at the top.
The compute power and data needed to train cutting-edge AI models are largely controlled by a few tech giants and elite labs. Researchers have noted a growing “compute divide”: modern AI research is “increasingly being shaped by a few actors…mostly affiliated with either large technology firms or elite universities”([Venture Beat](https://venturebeat.com/ai/ai-research-finds-a-compute-divide-concentrates-power-and-accelerates-inequality-in-the-era-of-deep-learning/#:~:text=“Exploiting the sudden rise of,”)) . This concentration is unsurprising when a single advanced model can require millions of dollars in cloud computing to train. In effect, those who control AI infrastructure can levy a de facto tax on progress – access to AI capability comes at a high price, like a new form of economic rent.
We find ourselves “taxed” by compute constraints: expensive GPUs, proprietary models, and limited cloud access act as toll gates on the road to AI-driven innovation.
These constraints raise a critical question for our AI age: will the fruits of automation be broadly accessible, or siphoned off by the few who can pay the toll? It’s a 21st-century twist on George’s dilemma. Just as Henry George witnessed railroads and factories enriching landowners while workers struggled. Compute power is the new land and railroads; data centers are the new factories.
Business owners could amass disproportionate advantage, widening inequality even as overall productivity soars, while reducing their workforces.
AI as the Great Equalizer
Despite these concerns, AI doesn’t have to repeat the past.
In fact, there are growing signs that artificial intelligence – especially generative AI – can become a great equalizer in society. How? By radically lowering the cost of creative and cognitive labor, AI is democratizing capabilities that were once the preserve of elites or large organizations.
Think of generative AI as a universal creative partner available to everyone. It can draft legal contracts, design logos, write software, or analyze data, often in seconds and at near-zero marginal cost.
Geopolitical strife complicates the picture further – nations race to lead in AI, stockpiling talent and semiconductors, sometimes restricting exports of critical chips or models.
This means that a solo entrepreneur today has access to “free intelligence” on tap – an army of cognitive assistants, essentially – which can level the field between them and big corporations. Bill Gates recently described this as entering a “[new era of ‘free intelligence](https://vjal.ai/bill-gates-within-10-years-ai-will-replace-many-doctors-and-teachers-humans-wont-be-needed-for-most-things/#:~:text=Bill Gates: Within 10 years,,be needed ‘for most things’)” powered by AI, one that will transform how humans work . When the cost of intelligence and creativity drops dramatically, abundance follows.
Routine blog posts, marketing copy, basic code – tasks that used to take teams of people and hefty budgets – can now be generated by anyone with an internet connection. The limiting factor is no longer manpower or money, but imagination and strategy in using these tools.
We already see small businesses and startups punching above their weight thanks to AI. A recent survey found that [44% of microbusiness owners believe generative AI will help them compete with larger firms](https://inbusinessphx.com/issuedate/may-8-2024#:~:text=Generative AI Levels the Playing,for Small Businesses, Says Survey). In other words, nearly half of the smallest entrepreneurs see AI as a leveling force, not a threat. And they’re acting on it: in that survey, 50% of those business owners had tried generative AI tools in just the past few months, tapping into everything from AI-assisted marketing to automated customer service.
Their optimism isn’t unfounded. As [Harvard Business Review](https://hbr.org/2024/02/genai-can-help-small-companies-level-the-playing-field#:~:text=Generative AI is arming smaller,the playing field more level) analysts observed, “Generative AI is arming smaller companies with once unattainable capabilities,” potentially making the playing field more level with big competitors.
When development cycles that once took months can be completed in days, and when going to market no longer requires a massive staff, the advantage of sheer size diminishes. In the words of one venture investor, “the democratization of powerful AI tools” is eroding traditional tech moats and shifting competitive advantage toward how effectively you execute and serve customers, rather than how many engineers you employ.
In this new landscape, good ideas can come from anywhere and succeed based on their merits, because access to advanced technology is no longer a guarded privilege.
AI is not just about efficiency – it’s a force multiplier for human creativity. A content creator or coder using AI can produce exponentially more output without additional human hours. Microsoft’s research found that developers using AI coding assistants (like GitHub’s Copilot) could complete tasks 55–60% faster on average ([UCSD.edu](https://blink.ucsd.edu/technology/about/news/posts/2024-08-01-github-copilot.html#:~:text=April ,orange) is with AI)).
That kind of productivity boost is transformative. It means a small team can iterate and innovate at a pace that only large companies could previously. It means a lone artist can generate concept art and animations that rival a studio’s output.
It even means individuals can personalize products or services at scale – think of teachers using AI to tailor lessons to each student, or doctors using AI to monitor and guide each patient. In short, AI is enabling personal-scale productivity that rivals industrial-scale productivity.
Every person gets a cognitive upgrade, akin to an employee who works tirelessly and with superhuman speed. This flips the script on the old economy of scale; bigger isn’t necessarily better when everyone has access to smart tools.
The result could be an explosion of creative and entrepreneurial activity across the globe. When the cost of producing something drops, we tend to get a lot more of it. (Economists would say the supply curve shifts dramatically right.)
We saw this with the internet lowering the cost of information distribution – suddenly there was an abundance of content and online businesses. We’re likely to see it again with AI lowering the cost of creation and problem-solving.
Imagine a future where a teenager in a rural village can design products using AI and sell them worldwide, or a small startup can disrupt an industry by deploying AI algorithms that only giant tech firms could afford a few years prior.
This isn’t wishful thinking; it’s already happening in embryonic form today. Open-source AI models are emerging that match the performance of corporate labs, but run on ordinary hardware. Meta’s Llama 2 model release in 2023 was a watershed moment – it demonstrated that [a cutting-edge language model could be optimized to run on consumer-grade computers, not just supercomputer clusters](https://medium.com/@ishanasabrish78/the-open-source-ai-revolution-how-llama-and-other-models-are-democratizing-artificial-intelligence-030a6e599915#:~:text=this revolution).
Suddenly, developers and researchers who couldn’t pay for unlimited cloud GPU time had a powerful model they could tinker with locally. A wave of innovation followed. Community-driven projects fine-tuned open models like [Llama, Falcon, and BLOOM](https://medium.com/@ishanasabrish78/the-open-source-ai-revolution-how-llama-and-other-models-are-democratizing-artificial-intelligence-030a6e599915#:~:text=While Llama may have captured,Here are some notable examples) for specialized tasks, from medical chatbots to educational tutors.
Small healthcare startups, for example, are now building AI assistants by training open models on medical literature – without needing enormous computing resources or exclusive data deals. Local businesses have begun deploying AI customer service agents fine-tuned to their niche, something that would have been prohibitively expensive if only proprietary AI were available.
This burgeoning open-source AI ecosystem is unlocking talent and creativity outside the big tech hubs, much the way open-source software (think Linux, or the web itself) did in earlier decades.
Like this article? Follow Eli Wood and Black Flag Design , give a listen to the AI DIY podcast! Your support and comments are appreciated.
