Nobody Wanted This Job, and Now It’s the Most Valuable Skill in AI

The article argues that data labeling and human-in-the-loop training are the hidden, valuable bottlenecks behind AI progress. It highlights Amazon’s Mechanical Turk as the original example of outsourced human labor powering machine intelligence, and notes that Amazon has stopped accepting new Mechanical Turk customers as of July 30. The piece then points to Scale AI as a major beneficiary of this trend, citing Meta’s $14.3 billion investment for a 49% stake and hiring of Alexandr Wang. Tesla is presented as another example, using its vehicle fleet and shadow mode to generate proprietary driving data at scale. The market takeaway is that companies controlling unique AI training data and labeling infrastructure may have durable moats, especially as robotics creates a new wave of data-collection needs. The article is largely thematic and promotional, with no direct price catalyst for Amazon but clear strategic relevance to AI infrastructure and data-moat businesses.