Report that ``OpenAI hired Kenyans at an hourly wage of 300 yen or less to improve ChatGPT''

OpenAI, an AI research organization, hires Kenyan workers at an hourly wage of $ 2 (about 260 yen) or less to make it safer to use interactive AI 'ChatGPT', and a huge amount used for data sets It is reported that it was working to filter harmful content from the text of.

????Exclusive: OpenAI used outsourced Kenyan workers earning less than $2 per hour to make ChatGPT less toxic, my investigation found

(Thread) https://t.co/302G0z7vy3

— Billy Perrigo (@billyperrigo) January 18, 2023

OpenAI Used Kenyan Workers Making $2 an Hour to Filter Traumatic Content from ChatGPT
https://www.vice.com/en/article/wxn3kw/openai-used-kenyan-workers-making-dollar2-an-hour-to-filter-traumatic-content-from-chatgpt

ChatGPT was taught by the world's poorest people | Metaverse Post
https://mpost.io/chatgpt-was-taught-by-the-worlds-poorest-people/

OpenAI has partnered with Sama , a company that undertakes tasks such as data labeling, and by detecting, labeling and eliminating harmful content from the data set that ChatGPT learns, ChatGPT is racist or sexist Try not to generate generic text.

Workers commissioned by Sama were responsible for sexual abuse, including child sexual abuse, incest, bestiality, murder, suicide, torture, and self-harm from ChatGPT's training dataset. It was filtering NSFW content that was violent or violent. As a result, TIME , an American weekly magazine that reported exclusively, points out that workers were forced to read the contents of these NSFW content.

Their working conditions reveal a darker side to the AI boom: that AI often relies on hidden, low-paid human workers who remain on the margins even as their work contributes to a multibillion-dollar industry. (6/8) pic.twitter. com/zLXfgGjIwM

— Billy Perrigo (@billyperrigo) January 18, 2023

A Kenyan worker who was actually in charge of the work said, ``I read a sentence about a man having sex with a dog while a young child was watching, and suffered from recurring hallucinations.That is torture. I will read a lot through it, ”he said. The salary was determined based on seniority and performance, and the hourly wage was 1.32 dollars (about 170 yen) to 2 dollars.

According to Motherboard, an IT-related news site, it is not limited to this example of ChatGPT, but it is often the case that AI innovation is supported by low-wage workers. Sama is working to filter images and videos of executions and child abuse from Meta's AI dataset, and it seems that this work was also done at an hourly wage of $ 1.5 (about 190 yen).

Immediately after TIME reported the details of Sama's work, Sama's CEO Wendy Gonzalez said that he had instructed the company to discontinue the work contracted from Meta and reduce the work of OpenAI. However, this decision caused workers who had contracted work from Sama to lose their jobs.

Motherboard said, ``AI experts want to highlight the human labor that builds the foundation of machine learning, focusing on how to ethically incorporate humans into processes rather than innovation. , acknowledging power imbalances, being more transparent about people in the loop , and improving working conditions to create opportunities for workers beyond data labeling and moderation. The fact that we're exploiting workers in building ChatGPT is a reminder of how far ChatGPT is from magic and glamour, and makes us reconsider how much we admire its innovation.'

However, regarding the TIME report that ``OpenAI is exploiting Kenyan workers'', the hourly wage in Kenya is $ 2, that is, the monthly salary is $ 320 (about 41,000 yen) than the average living wage in Kenya It is also pointed out that the workers are being paid enough because the wage rate is 53% higher.

$2/hour = $16 per 8 hour day = ~$320/week, which is 53% above the Living Wage in Kenya https://t.co/DX7SWSxg5s pic.twitter.com/3C17IwV3cg

— Steven Cohn (@spcohn) January 18, 2023