Mom Balances Real Estate and AI Side Hustle, Earning $8,000 in Just Three Weeks

The Life of a Part-Time AI Trainer: A Closer Look at Amanda Overcash
A Day in the Life of Amanda Overcash
Amanda Overcash, a dedicated single mother based in Texas, manages two demanding roles in her life. By day, she navigates the world of real estate, but as night falls and her daughter goes to bed, she transforms into an AI trainer. With her headphones in and pajamas on, she sits at her kitchen table, working tirelessly on her second job.
Many nights find Overcash working until midnight, while other times she rises at 4 a.m. to squeeze in an extra hour of work before her full-time job begins. Her dual roles highlight the often-hidden labor force that supports the rapid advances in artificial intelligence.
The AI Data Industry
Overcash is one of many individuals contributing to the AI boom, a largely unrecognized yet crucial segment of the workforce. While some individuals in this field share negative experiences, Overcash has largely had a positive journey, drawn in by potential earnings that can reach up to $40 per hour. Last summer, in just under three weeks, she earned nearly $8,000 from her work, largely related to writing and rating chatbot responses.
Despite the promise of high wages, Overcash emphasizes that this line of work is not a get-rich-quick scheme. Tasks can be demanding, and the mental toll of balancing both a full-time job and a side hustle can lead to burnout.
Diversified Projects Across Platforms
In her six years in the AI data industry, Overcash has engaged in various projects such as transcription, ad moderation, and evaluating prompts. Many individuals like her work across multiple platforms to maximize opportunities. Notable platforms offering such jobs include Appen, OneForma, Prolific, and Amazon Mechanical Turk. Appen, for instance, boasts a diverse workforce of over one million contractors across 200 countries.
The tasks can vary dramatically, encompassing everything from labeling satellite imagery and transcribing recordings to reviewing chatbot outputs and even uploading pet videos. Pay rates depend on the complexity of the job; Overcash points out that tasks related to large language models (LLMs), which drive generative AI, typically command higher wages of about $20 per hour, whereas social media-related projects might pay between $9 and $11.
Tasks and Responsibilities
Currently, Overcash splits her time between two main projects. One involves transcribing casual voice memos, which often capture noise from bustling environments like cafés or busy kitchens.
“They are supposed to sound natural, but it can be challenging due to background disturbances,” she explains.
In addition to transcription, Overcash also evaluates social media advertisements. She reviews each ad, considering various factors such as the presence of nudity or profanity, the accuracy of claims, and age appropriateness. This task is more straightforward for her and is one that she enjoys because it requires minimal second-guessing.
High Earnings Potential
Overcash’s earnings demonstrate the financial potential of these roles. Recently, she worked intensely on a chatbot evaluation project, putting in up to 16 hours a day. Starting at $22 per hour, her pay rose to $40 as she became more entrenched in the project, ultimately accumulating nearly $8,000 in less than three weeks.
Her responsibilities included reviewing chatbot responses to sensitive health questions and determining if any advice could be harmful. “For instance, if a user asking about a lump didn’t receive advice to seek medical help, I had to mark that as unsafe,” she noted.
Challenges and Realities of AI Work
Despite the allure of high pay, Overcash stresses that the work is demanding. Tasks come with the pressure of audits and passing rigorous tests to qualify for projects. Onboarding to platforms is a challenge in itself; many require thorough understanding through assessments based on detailed guidelines.
“Getting on a project is tough. Some tests took me days to finish,” she stated, highlighting the intensity of the initial process.
Once she is part of a platform, the worry doesn’t end. Regular audits often occur without forewarning, and a single failure in these audits can lead to losing access to assignments.
Finding Balance in Life and Work
After experiencing burnout two years ago, Overcash adjusted her approach to her AI side hustle. She has set clear boundaries, including a strict rule to avoid working on weekends to dedicate time to her daughter. Flexibility is key; she may work two hours on some days and up to eight on others, balancing her responsibilities with care.
While Overcash has mostly encountered rewarding projects, she acknowledges that the industry can sometimes be unpredictable, particularly concerning issues faced by some platforms.
Despite the challenges, Overcash finds value in her work. The various tasks have sharpened her analytical skills and made her more attuned to identifying biases in language. While the work doesn’t come without stress, it offers her the necessary flexibility, consistent income, and control over her schedule.