Search
Content type: Examples
Human raters have played a significant role in the rapid improvement in the machine learning models that fuel modern AI. The raters evaluate the algorithmic output of search engines and AI chatbots and provide "Reinforcement Learning with Human Feedback" (RLHF) – the technical name for the deployment of such ratings to improve AI models. The efforts of these workers, who are mostly located in the global South but include thousands in the US, is downplayed by the technology companies to whom…
Content type: Examples
Delivery drivers in Jakarta use GPS-spoofing apps in order to improve their chances of selection by the Gojek delivery and transport app, an equivalent to Apple Pay, Postmates, Venmo, and Uber all in one. Gojek that operates in more than 200 cities in Indonesia, Singapore, Vietnam, and Thailand. Other grey market apps enlarge details of orders that are too small to read, automate bidding, and apply filters to open orders. Some apps are distributed via Google Play; more are sold via driver…
Content type: Examples
Four people in Kenya have filed a petition calling on the government to investigate conditions for contractors reviewing the content used to train large language models such as OpenAI's ChatGPT. They allege that these are exploitative and have left some former contractors traumatized. The petition relates to a contract between OpenAI and data annotation services company Sama. Content moderation is necessary because LLM algorithms must be trained to recognise prompts that would generate harmful…