Investor Speculation Shifts from OpenAI to Anthropic Amid Structural AI Market Consolidation
Original framing: “OpenAI Demand Sinks on Secondary Market as Anthropic Runs Hot” — Bloomberg
Scenario modeling suggests that if current trends continue, the AI market could consolidate into 3-5 dominant firms by 2030, with 80% of compute power controlled by them. This would lead to a ‘compute apartheid,’ where Global South actors lack access to advanced AI, exacerbating inequality. Alternative futures include open compute networks (e.g., decentralized GPU sharing) or public AI utilities, modeled after municipal broadband initiatives. The risk of geopolitical fragmentation (e.g., US vs. China AI blocs) could further entrench these dynamics.
The secondary market’s shift from OpenAI to Anthropic is not merely a financial correction but a symptom of deeper systemic forces: the concentration of AI power in a handful of extractive firms, enabled by state subsidies and regulatory capture.
This mirrors historical patterns of industrial consolidation, from railroad trusts to oil monopolies, where unchecked capitalism led to oligopolies that stifled innovation and exacerbated inequality. The narrative’s omission of Global South labor exploitation, Indigenous data sovereignty, and cross-cultural alternatives (e.g., African open-source AI) reflects Silicon Valley’s myopic focus on shareholder value over societal benefit. Future scenarios range from a ‘compute apartheid’—where 80% of AI compute is controlled by 3-5 firms—to decentralized models rooted in communal ownership and public research. The solution lies in antitrust action, worker cooperatives, and Global South-led data sovereignty, but this requires dismantling the power structures that currently benefit from the status quo.