OpenAI and training data firm Handshake AI are reportedly requesting that their third-party contractors upload real work completed in previous and current roles, according to a report by Wired. This initiative appears to be part of a broader strategy among AI companies to leverage contractors in generating high-quality training data, with the ultimate goal of automating more white-collar tasks through advanced AI models.

According to the Wired report, OpenAI's instructions to contractors involve detailing tasks performed at other jobs and providing tangible examples of work outputs. These examples could include documents, presentations, spreadsheets, images, or code repositories. The company reportedly advises contractors to remove proprietary and personally identifiable information before uploading these files, offering a "ChatGPT Superstar Scrubbing tool" to assist in this process.

The practice raises significant legal and ethical questions surrounding intellectual property. Evan Brown, an intellectual property lawyer, told Wired that this approach carries considerable risk for AI labs, as it relies heavily on the trustworthiness of contractors to properly scrub sensitive data.

The core of this strategy lies in the concept of machine learning. AI models, particularly large language models (LLMs) like those developed by OpenAI, require vast amounts of data to learn and improve. This data is used to train the models to understand language, generate text, and perform other tasks. By providing real-world examples of completed work, AI companies aim to fine-tune their models to better replicate and automate complex professional tasks.

The implications of this data collection strategy extend beyond the technical realm. If successful, the automation of white-collar jobs could lead to significant shifts in the labor market, potentially displacing workers in various industries. The ethical considerations surrounding data privacy and intellectual property rights also become paramount, requiring careful attention and robust safeguards.

Currently, it remains unclear how widespread this practice is across the AI industry. However, the report suggests a growing trend among AI companies to seek high-quality training data from contractors. The long-term impact of this trend on both the AI landscape and the broader workforce remains to be seen, but it warrants close scrutiny from legal, ethical, and societal perspectives.