The tech industry wrongly compares AI to software, which has near-zero marginal costs for new users. In reality, providing access to frontier AI models is a zero-sum game during compute crunches because of immense computational requirements. Servicing another user is expensive, leading to rationed access.

The current need for massive data and compute, which concentrates power in large companies, is a characteristic of the Transformer architecture, not an inherent law of AI. Future research breakthroughs will likely enable smaller, more efficient models, decentralizing AI development and power.

Unlike social media, which scaled without physical impediments, AI's progress depends on massive, resource-intensive data centers. This physical footprint makes the industry vulnerable to local political opposition, regulations, and even violence, creating a new bottleneck for growth that pure software companies never faced.

New communication technologies historically democratize society before elites regain control. Jill Lepore argues that emerging AI technologies are breaking this cycle, as they appear to be consolidating power with elites from the outset rather than providing an initial wave of popular empowerment.

Decentralized power has been a key driver of capitalist growth. However, AI exhibits immense economies of scale in training, data, and R&D. This suggests a future where hyper-centralized "AI economies" within a few firms could grow much faster than the broader, decentralized market, inverting a core economic principle.

Contrary to the idea of AI for all, the most powerful models will likely be restricted to a few high-paying clients to prevent distillation and maximize revenue. This creates a future where competitive advantage is defined by exclusive AI access, potentially allowing large incumbents to crush smaller competitors.

Unlike previous tech waves, AI's core requirements—massive datasets, capital for compute, and vast distribution—are already controlled by today's largest tech companies. This gives incumbents a powerful advantage, making AI a technology that could sustain their dominance rather than disrupt them.

Massive capital accumulation resulting from AI investments creates systemic democratic risks beyond simple unfairness. Lovely points to tech billionaires leveraging AI gains to acquire major media conglomerates like CBS, Paramount, and the Washington Post. When ultra-wealthy owners control critical public discourse, they can compromise journalistic independence to protect core corporate interests or curry favor with political administrations.

Meredith Whittaker argues the biggest AI threat is not a sci-fi apocalypse, but the consolidation of power. AI's core requirements—massive data, computing infrastructure, and distribution channels—are controlled by a handful of established tech giants, further entrenching their dominance.