#OpenAI has exploited third-world workers for less than $2 an hour to make its “Artificial Intelligence” #ChatGPT less toxic.
AI from OpenAI owes its capabilities to the huge amounts of data taken from the Internet, from which it has also absorbed some of the prejudices (racism, sexism, etc.) that make the software difficult to sell. To remedy this, the company has begun to build a security mechanism based on artificial intelligence that recognizes toxic sentences and eliminates them, relying on #Sama, a Californian company already at the center of scandals between trauma and denied rights that outsources labelers in third world countries with miserly salaries.
The work of labeling is a very exhausting work from a psychological point of view, since it is necessary to read all day, every day, texts that may contain disturbing and traumatizing content.
The collaboration between OpenAI and Sama collapsed in February when the latter refused to label images that could contain violence, child abuse and other disturbing material.
https://time.com/6247678/openai-chatgpt-kenya-workers/
AI from OpenAI owes its capabilities to the huge amounts of data taken from the Internet, from which it has also absorbed some of the prejudices (racism, sexism, etc.) that make the software difficult to sell. To remedy this, the company has begun to build a security mechanism based on artificial intelligence that recognizes toxic sentences and eliminates them, relying on #Sama, a Californian company already at the center of scandals between trauma and denied rights that outsources labelers in third world countries with miserly salaries.
The work of labeling is a very exhausting work from a psychological point of view, since it is necessary to read all day, every day, texts that may contain disturbing and traumatizing content.
The collaboration between OpenAI and Sama collapsed in February when the latter refused to label images that could contain violence, child abuse and other disturbing material.
https://time.com/6247678/openai-chatgpt-kenya-workers/
TIME
Exclusive: OpenAI Used Kenyan Workers on Less Than $2 Per Hour to Make ChatGPT Less Toxic
A TIME investigation reveals the difficult conditions faced by the workers who made ChatGPT possible
#Kenya: #Sama, a Kenyan company that handles #Facebook #content #moderation to reach production quotas, exposes its employees to disturbing material with rhythms that lead to the development of trauma: the mandatory weekly "wellness break" must be begged for, you have 50 seconds to moderate individual contents (including videos of minutes) maintaining an accuracy of 84%, you are monitored during the work activity and there is no trace of the act of renouncing for possible trauma (present in EU countries). In addition, when asked to raise salaries - Facebook pays them about $2 an hour - the company responds by saying that it does not want to upset the local market.
When two years ago an internal union tried to claim its rights, the company dismantled it, promising promotions and firing its promoter. He continues to suffer from disorders related to his former job, without at the same time being able to afford a therapist.
"Facebook is enriching itself by exporting trauma along old colonial axes of power, away from the United States and Europe, to the developing world".
https://time.com/6147458/facebook-africa-content-moderation-employee-treatment/
When two years ago an internal union tried to claim its rights, the company dismantled it, promising promotions and firing its promoter. He continues to suffer from disorders related to his former job, without at the same time being able to afford a therapist.
"Facebook is enriching itself by exporting trauma along old colonial axes of power, away from the United States and Europe, to the developing world".
https://time.com/6147458/facebook-africa-content-moderation-employee-treatment/
Time
Inside Facebook's African Sweatshop
Workers in the Nairobi office are among the lowest-paid workers for the platform anywhere in the world