What Are We Feeding AI? 

A working mother took a year-long break from her full-time job as a flight attendant to stay home after her second child’s birth. Chelsea Becker discovered her side hustle through TikTok videos on a website called Data Annotation Tech. This platform trains artificial intelligence models using provided information.  

Ms. Becker spent a few hours each day interacting with the AI-powered chatbot on Data Annotation Tech. She earned $20 to $40 for every hour worked. By the end of December through March, it was estimated that she was earning over $10,000.  

And It Works Like This  

It’s common knowledge that A.I. models derive knowledge from human input. Over the years, creators of A.I. systems such as Google and OpenAI have relied on underpaid workers, often contractors from external firms, to assist computers in visual recognition tasks. (The New York Times has filed a lawsuit against OpenAI and its collaborator, Microsoft, alleging copyright infringement.) These workers may annotate images for autonomous vehicles or identify objects in photos used to train A.I. algorithms. 

However, as A.I. technology advances, so does the complexity of the tasks assigned to human trainers. What used to be a simple job of tagging photos has evolved into more sophisticated tasks, including writing essays. 

Typically, there are two main types of tasks assigned to these trainers: supervised learning, where A.I. learns from human-written content, and reinforcement learning, where chatbots learn from human feedback on their responses. 

The Issue in This Case  

In this scenario, the concern lies not in the job opportunities created by feeding AI, but rather in the quality of information that is being fed to it on various websites like Data Annotation Tech. Consider Chelsea Becker, who provided the AI chatbot with parenting tips. How can one make sure that the advice she has given is globally applicable and valid? It’s easy to overlook the fact that people worldwide access various chatbots, leading to potential cultural discrepancies in the effectiveness of parenting strategies. 

Moreover, this situation raises the risk of parents unconsciously comparing their parenting methods to Chelsea Becker’s, potentially creating an identity crisis. While feeding AI on Data Annotation Tech is important for its development and assistance in daily tasks, it’s vital to acknowledge the potential downfall linked with the information provided to it. Let’s remain cautious about the data AI is being fed in order to avoid multiple issues along the way.  

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.