A clear hierarchy exists in AI development, with Anthropic (best talent, compute efficiency) and OpenAI at the top, Google at risk of falling out of the top tier, and Meta/xAI lagging due to talent and execution deficits.
The US government, particularly the Department of War, is aggressively pursuing unaligned AI for military and surveillance purposes, creating direct conflict with safety-oriented labs like Anthropic.
The current AI era is the 'middle game'; the 'endgame' will begin only when AI development becomes autonomous and human researchers are no longer the bottleneck for progress.
China is incapable of domestically manufacturing competitive AI chips at scale within a five-year timeframe, effectively sidelining it in the current hardware-dependent AI race.
AI's economic disruption is unprecedented because it will automate the new jobs it creates, preventing a traditional re-employment equilibrium and making stock ownership a flawed long-term wealth strategy.
▶The AI Lab Competitive HierarchyApr 2026
Moschewicz posits a clear, tiered structure for AI labs. He places Anthropic and OpenAI at the top, citing Anthropic's superior talent and compute efficiency, while viewing Google as being in 'danger of dropping out.' He categorizes Meta and xAI as a lower tier that is 'falling farther and farther behind' due to talent deficits and leadership issues.
Investors should focus on talent acquisition and compute efficiency as key performance indicators for AI labs, as these factors, more than just raw compute, currently determine leadership in the field.
▶Government Conflict and AI MilitarizationApr 2026
A central theme is the tension between AI developers and US government entities. Moschewicz highlights the Department of War's memo to advance AI 'even if it is not aligned' and its push for 'all-lawful use' contracts, which clashed with Anthropic's refusal to permit use for domestic mass surveillance.
The escalating demands from government and military actors represent a significant operational and ethical risk for leading AI companies, potentially forcing them to choose between lucrative contracts and their stated safety principles.
▶Defining the AI 'Endgame'Apr 2026
Moschewicz distinguishes between the current 'beginning of the middle game' and a future 'endgame.' He defines the endgame as the point when humans are no longer in control of the AI development loop and a lab's progress becomes a direct function of its compute, rendering human research talent irrelevant.
Analysts should monitor for signs of AI-driven research autonomy as the key indicator of a paradigm shift, which would dramatically alter the competitive landscape and devalue human capital as a primary asset for AI labs.
▶Unprecedented Economic DisruptionApr 2026
Moschewicz argues that AI's economic impact differs fundamentally from past technological waves. He claims that while previous automation created new jobs for humans, AI will also be capable of performing these new jobs, preventing a re-employment equilibrium and making traditional wealth strategies like stock ownership flawed.
This perspective suggests that the long-term economic impact of AI may not follow historical patterns, requiring a reassessment of labor market forecasts and long-term investment strategies that rely on sustained human economic productivity.