Title: ChatGPT’s Impact on Crowd Work: Challenges and Controversies
Crowd work platforms like Amazon’s Mechanical Turk have become a crucial source for researchers studying human behavior. Workers, such as Bob (name changed for anonymity), spend hours completing surveys and simple digital tasks in exchange for modest sums of money. However, the repetitive nature of the work, especially when answering questions about negative emotions, can become monotonous. This is where AI language models like ChatGPT step in to offer some relief.
Bob admits to using ChatGPT occasionally to generate responses that fit the desired mood set by the questionnaires. As someone who relies on crowd work for income, efficiency is paramount. Bob, like many others, opts for ChatGPT’s assistance to complete tasks more quickly. While crowd workers are often criticized for using AI language models as shortcuts, it is worth acknowledging the challenges and precarious situations they face.
The growing popularity of generative AI tools like ChatGPT extends beyond crowd workers to encompass various professions including students, office workers, coders, and even dungeon masters. These tools have received both praise and suspicion due to their potential to optimize work processes. As a result, some platforms are now implementing policies or technology to discourage or detect the use of AI language models like ChatGPT. However, it is crucial to strike a balance so as not to unfairly burden workers who already face precarious conditions.
A recent preprint study conducted by academics from the Swiss Federal Institute of Technology gained significant attention when it suggested that more than a third of Mechanical Turk workers had utilized ChatGPT to complete a text summarization task designed to measure human understanding. However, this claim was met with pushback from workers and researchers who defended the integrity of crowd workers and suggested that clearer instructions could help mitigate the issue.
In response to these concerns, CloudResearch, a company facilitating online studies, conducted their own study and found that only around 20% of their pre-screened workers used ChatGPT. Interestingly, when asked explicitly not to use AI, almost no participants relied on it. This indicates that given appropriate guidance and instructions, crowd workers are unlikely to resort to AI language models.
One crowd worker in her fifties, who actively engages within an online community of Turkers, asserts that the majority of workers would never consider cheating. According to her, integrity is a core value for crowd workers, many of whom prefer this type of work setup for its flexibility and autonomy, which can be particularly beneficial for introverts or neurodivergent individuals. Another worker, who relies on Mechanical Turk due to an illness that confines her to working from home, expresses a similar sentiment, stating that risking account suspension would jeopardize her livelihood.
In conclusion, the use of AI language models like ChatGPT among crowd workers has raised concerns regarding the integrity of their work. However, it is important to acknowledge that the majority of crowd workers are motivated by a sense of integrity and take pride in their contributions. As platforms implement measures to discourage AI usage, it is crucial to strike a delicate balance that does not unfairly burden workers. Clearer instructions and ongoing discussions within the crowd work community can help address these challenges and safeguard the integrity of this essential field that allows individuals to work on their own terms.
