Key facts
- AI companies are seeking work-related data to train models for automating tasks.
- Alternative data providers aim to supply these datasets to AI companies.
- The market for alternative data experienced growth in the mid-2010s.
Artificial intelligence companies are actively seeking large volumes of work-related data, including internal memos, transaction records, and contracts, to train their AI models. This demand stems from the need for AI tools to deeply understand work processes to automate them effectively. This push for data could revitalize the market for alternative data, which consists of unstructured datasets like satellite imagery and credit card transactions. This sector saw a surge in interest on Wall Street in the mid-2010s, as providers aimed to offer investors an edge through non-traditional data. While some new startups are emerging to package data for AI models, challenges related to scaling offerings and maintaining data quality, which plagued the alternative data market previously, may persist.