Are we training AI to take our own jobs? It sounds like the premise of a science fiction drama, but for a growing number of experienced professionals it has become a genuine question. Writers, directors and producers are now being paid to teach artificial intelligence how to perform tasks that once depended entirely on their knowledge and experience.
The work can pay well, particularly for people whose regular employment has become uncertain. However, it creates an uncomfortable contradiction. Someone may receive a welcome payment today for improving a system that could reduce demand for their skills tomorrow.
Hollywood Professionals Are Teaching AI Their Craft
A recent Guardian investigation found that experienced Hollywood creatives are accepting temporary work training AI models. Their assignments include teaching systems how to develop screenplays, prepare production schedules, assess filming requirements and assemble pitch presentations.
These are not simple administrative exercises. Planning a film shoot may require an understanding of permits, locations, lighting, safety, staffing, budgets and the individual needs of performers. Much of that knowledge is normally acquired through years of practical experience. AI companies need knowledgeable people because a model cannot learn professional judgement merely by reading a collection of scripts or production manuals.
Payments reportedly range from $12 to $200 an hour, depending on the work and the person’s expertise. For professionals struggling to find conventional employment, refusing that income is not always a realistic option. The decline in film and television work has left many talented people facing a difficult choice between immediate financial security and concern about what their contribution might eventually enable.
One participant described the experience as being handed a shovel to dig the grave of their own profession. That image is dramatic, but it captures the anxiety surrounding this new form of work.
Why AI Still Needs Experienced People
Artificial intelligence models learn from examples, corrections and evaluations. They may produce an answer, plan or creative draft, but somebody must decide whether it is convincing, accurate and professionally useful. That requires people who understand the work at a much deeper level than the model initially does.
This is why the story reaches far beyond Hollywood. AI developers recruit specialists from finance, law, healthcare, education and social work. A qualified professional may be asked to design realistic problems, assess the model’s answers or demonstrate how an expert would approach a difficult case.
The most valuable material is often not a fact that could be found online. It is the reasoning behind a decision. By explaining that reasoning, workers are transferring part of their professional judgement into a system that can reproduce elements of it at enormous scale.
A New Way of Capturing Workplace Knowledge
Businesses have always recorded knowledge through manuals, training courses and documented procedures. The difference is that conventional material helps another person perform a job. AI training can help a machine imitate parts of the worker’s decision making.
This does not mean that an entire occupation can suddenly be automated. Most jobs contain a mixture of technical tasks, relationships, accountability and unpredictable situations. Nevertheless, if AI can perform enough routine or preparatory work, an employer may decide that fewer people are required.
The immediate risk may therefore be the gradual removal of tasks rather than the disappearance of complete professions. This matters because junior tasks are also how people learn. If entry level work is automated, organisations may eventually struggle to develop the experienced professionals they still need.
Why Workers Agree to Participate
It is easy to criticise someone for helping to train a potential competitor, but that ignores the economic reality. Freelancers and creative professionals cannot pay their bills with a principled refusal. If traditional work is already declining, AI training may be one of the few opportunities available to them.
There is also no guarantee that refusing would prevent the technology from developing. Another person may accept the assignment, or the company may use less suitable data. Some workers therefore see participation as a practical response to an unavoidable change.
The ethical burden should not be placed entirely upon individuals. AI companies and employers decide how these systems are deployed. Governments determine what protections workers receive. A person accepting temporary work during a downturn is not responsible for the wider structure of the AI economy.
AI Could Also Expand Creative Opportunity
There is another side to the argument. Technology that reduces production costs can allow people to create work that would never have attracted conventional funding. Independent filmmakers may use AI for visual effects, planning, translation or editing and produce ambitious projects with very small teams.
AI does not have to replace professional creativity. It can act as a tool that enables a skilled person to attempt more. The important question is who receives the benefit. If productivity gains are shared through better pay, shorter working hours and new opportunities, the technology could improve working life. If they are used only to reduce headcounts and concentrate profits, resistance will understandably grow.
Businesses Need an Honest Plan for the Workforce
Organisations introducing AI should be clear about what they are trying to achieve. Employees deserve more than vague assurances that the technology will simply assist them, particularly when systems are being trained upon their own work.
Transparency matters because workers cannot make informed decisions if they do not know how their contributions will be used. Contracts for AI training should explain whether submitted material may become a permanent part of a commercial product, how confidential information will be protected and whether the trainer retains any rights over original work.
The Real Issue Is Who Controls the Transition
The Hollywood story provides an unusually clear picture of a much wider change. Human expertise is becoming one of the essential raw materials of artificial intelligence. The people supplying it may benefit from new income and opportunities, but they may also help companies automate parts of the careers they spent years developing.
We should not conclude that every AI trainer is eliminating their own job. Technology rarely changes employment in such a simple or immediate way. It removes some tasks, alters others and creates opportunities that were previously difficult to imagine.
However, the concern is legitimate. Workers should not be expected to train new systems in silence and then discover that the resulting productivity gains have been used against them. The future of work will not be determined by AI alone. It will depend upon the choices made by businesses, policymakers and society about how the benefits are divided.
The question is therefore larger than whether people are training AI to take their jobs. It is whether they will have any meaningful voice in deciding what happens once the training is complete.
Sources
The Guardian, “Digging the Grave of My Profession: The Hollywood Creatives Training AI to Do Their Jobs”, 22 August 2026.
FilmLA Research, Los Angeles production reports.
US Bureau of Labor Statistics, motion picture and sound recording employment data.
Explore More Than 110 AI Courses
Explore more than 110 AI courses at AI Tuition Hub, covering everything from complete beginner lessons to AI in business, finance, education and the workplace. Access every course for only £9.95 a month.
Published: 22nd August 2026.