I argue that the current model does not take into account how data classification projects differ significantly from those in other computer science fields. For instance, in programming, making mistakes is a part of the learning process for programmers, which helps them gain experience and avoid making the same mistakes in future projects.

In contrast, each annotation project has unique criteria and goals, and the time and effort invested in figuring out what the training guidelines have omitted does not transfer to the next project. Due to poor training methods we make mistakes and have to redo our own work until we figure out all project details. By the time we have figured the process out, the project is often already completed, leaving us to start anew with a different project, and the cycle begins again.

I believe a well-structured training process should value our time and hard work. After discussing with my co-workers and compiling a list of issues, it became clear that many problems stemmed from the training itself. For example, my colleague Moaiad expressed his frustration over a significant gap between what was promised and the reality; he was assured a five-hour workday but ended up spending an additional five hours correcting mistakes from the first five. Compensation is a further source of frustration.Another colleague, Hind, mentioned that she invested heavily in internet and electricity for her work, hoping it would yield a modest return. Regrettably, compensation was always subject to change at the end of the project and depended on her speed and error rates, which left her with minimal profits.

Convinced that a short film can have a greater impact than academic writing, I decided to create a documentary that captures my everyday experiences as a Syrian data worker. My hope is that this visual narrative will prompt those involved in the AI industry to recognize the true nature of our daily labor. My personal experience and informal conversations and interviews conducted with colleagues are the basis for this account and the reason why I focus on the challenging circumstances we face at work and the flaws in our training system. 

I call for a structured and fair training program, recognition of our daily struggles, and a commitment to changes that value our contributions as integral to the AI ecosystem. I also demand a supportive work environment that empowers us, the data workers, and strengthens the AI industry to benefit its workers, not just the companies employing them.