Tech giants like Google, Amazon, and Microsoft are investing well in the AI race , and to develop generative AI tools, which generate answers to textual prompts in the form of text and images, and these companies employ thousands of invisible contract workers “subemployees” who train and fix the algorithms. And optimize them to deliver the final product to end users in seconds.
According to the report, these “coaches” are underpaid, overworked, and overworked by companies that have invested billions in developing their AI model.
According to a report by Bloomberg, US lawmakers have asked big tech companies to disclose details about their “ghost business.”
This task force consists of invisible workers who label data and evaluate responses, and they also play a pivotal role in training the AI.
“Although... "It is an essential nature of this work, but millions of data workers around the world perform these stressful tasks under constant supervision, for low wages and no benefits."
Workers are expected to screen dangerous chatbot answers, but they may have little time to evaluate the integrity of answers, and data workers often have little training or supervision, which can lead to bias.
Lawmakers also required companies to answer questions about their data workforce, including their ability to take breaks, appeal, or access mental health resources due to distressing content.
Aside from the mentioned companies, other AI-focused companies like OpenAI, maker of ChatGPT, Inflection AI, Scale AI, and Google-backed Anthropic also received the letter.
The report claims that to develop AI products, US companies rely heavily on subcontracted employees.
These workers are located either locally or abroad and are recruited through external staffing companies and these “ghost” employees often lack the benefits provided to direct employees of the companies.
These companies also rely on such work for other tedious tasks such as content moderation and ensuring product quality.
In January, Time magazine reported that Microsoft-backed OpenAI paid workers in Kenya less than $2 an hour to filter out questionable content from ChatGPT.
Google ChatGPT's competitor, Bard, also relies on human labor and the report claims that these workers have little training, high workloads and low wages.