The Great Decoupling of Labor and Capital

The post argues that megacap tech companies have been adding revenue with far fewer employees than in earlier decades. It compares Apple, Alphabet, Microsoft, Meta, Amazon, Nvidia, and Walmart to show that incremental revenue has required progressively less incremental headcount, even before generative AI.

It then extends that pattern to AI and market structure. The author says AGI is better understood as an asymptote than a single event, revises upward expectations for OpenAI’s revenue path, and suggests the broader economy may look more like a smiling curve, with big platforms and many smaller participants benefiting while the middle gets squeezed.

Reading notes

#

• HP and IBM crossed the $100 billion revenue mark with very different employee counts, and the post uses that contrast to frame a long-run decoupling of revenue growth from labor growth.
• Apple, Alphabet, Microsoft, and Meta all reached later revenue milestones with much smaller additions to headcount than they needed for their first $100 billion.
• Amazon stands out as a partial exception, but the author still argues that its most recent $200 billion in revenue came with relatively limited incremental hiring.
• Nvidia and Walmart are used to show that the pattern is not limited to pure tech companies.
• The author summarizes the pattern as roughly $1 trillion of incremental revenue coming with only about 100,000 new jobs.
• The post argues that AGI is not likely to arrive as a single corporate announcement, but as a gradual asymptote that AI helps speed up.
• The author revises expectations for OpenAI, saying $200 billion of revenue by 2030 looks less remote than before.
• The post says investors often frame big tech as direct rivals, but that this can miss how new products expand markets and leave consumer surplus.
• The broader economy is described as a smiling curve, where dominant platforms and many smaller creators or sellers benefit while middle layers are squeezed.