The Dual Edge of AI: Progress and Inequality
Instead, I saw a real aristocracy, armed with a perfected science and working to a logical conclusion the industrial system of to-day. Its triumph had not been simply a triumph over Nature, but a triumph over Nature and the fellow man.
— H. G. Wells, The Time Machine, 1895
We find ourselves at a tricky crossroads: how do we allow companies to operate efficiently without allowing them to become monopolies? On one side, efficiency is a key driver of innovation and lower prices, but on the other, unchecked corporate power can stifle competition and widen economic divides. Economists like Jan Eeckhout have warned us about this. In his work, The Profit Paradox, Eeckhout describes how large firms have become so powerful that they suppress wages and keep prices artificially high, limiting opportunities for smaller companies and driving inequality.
In the midst of this tension, we need to make sure the benefits of technological advances like AI are evenly distributed to all workers, not just the select few in high-skill, high-wage industries. We’re already seeing that AI productivity gains are mainly benefiting high-income workers while leaving behind those in manual or service jobs. This disparity could grow if we don’t make deliberate choices about how AI is integrated into different sectors of the economy.
This is especially important in developing countries, where the gap between technological haves and have-nots is even more pronounced. But there’s hope. As the cost of computing power decreases, access to AI should become more widespread. This expansion of AI could help reduce some of the bias that stems from data only representing wealthier, more developed regions. In countries like Uganda and India, AI is already being used to close critical gaps in education and public services. You can see this trend at work in projects like AI-driven education tools that help students in underfunded schools or AI systems that streamline public services.
That said, if we want AI to have a truly equitable impact, we’ll need to regulate it. Just like how telecoms were regulated to allow multiple providers on one network, we need similar regulation for AI. This would give smaller companies access to the infrastructure they need to compete with the tech giants. Economists like Eeckhout have proposed relaxing intellectual property protections so that patents don’t give firms like ASML a permanent stranglehold on the market. It’s about creating an environment where innovation isn’t just possible for the biggest players.
Another area that deserves more attention is how AI can serve low-skilled workers. There’s a tendency to focus on white-collar workers and their interaction with AI, but what about manual AI tasks? These are jobs that don’t require a computer but still involve interacting with intelligent systems in ways that can augment productivity. Focusing on reskilling low-skilled workers for these roles could provide more opportunities in a world increasingly dominated by automation.
Speaking of reskilling, the future of AI in education is another frontier we can’t afford to ignore. Imagine the possibilities of AI co-piloting in the classroom. Tools like Khanmigo from Khan Academy are already proving this concept. These AI-driven systems allow teachers to personalize learning experiences, while AI handles some of the more repetitive or administrative tasks. This frees up teachers to focus on what they do best — teaching and guiding students through complex subjects.
Ultimately, we stand at a pivotal moment where the choices we make about AI could either reinforce inequality or help dismantle it. The technology itself is neither good nor bad — it’s all about how we choose to use it. By ensuring that AI’s benefits are shared more broadly, regulating the tech giants, and investing in education, we can build a future where progress doesn’t come at the expense of fairness.
Thanks for reading! I’m always looking for new topics around AI to write about, please message me to request a topic!
Robert
