The VFX industry watched computers become affordable, software become accessible, and training become available almost everywhere. From this genuine progress, it drew a dangerous conclusion: artists could now create themselves.
They cannot.
Tools can be acquired. Skills can be taught. Professional judgment still has to pass from one generation to the next, and that is the part of the system the industry is steadily dismantling.
Studios continue to complain that experienced artists are difficult to find. Yet across much of VFX, junior positions have narrowed, contracts have shortened, crews have become less stable, schedules have tightened, and senior artists have been given less time to develop the people working beside them. The industry still expects judgment, speed, visual intelligence, technical discipline, pipeline awareness, and calm under pressure. What it increasingly refuses to support is the environment in which those qualities are formed.
A sustainable path into VFX rests on three pillars: access to tools and the ability to practise, structured training, and professional mentorship inside real production.
Access makes sustained practice possible. Training gives that practice direction, principles, criticism, and technical structure. Professional mentorship connects knowledge to consequence. It teaches artists how their decisions behave when other people, departments, schedules, budgets, and clients depend on them.
These pillars support one another, but they are not interchangeable. Access without training can produce years of unfocused experimentation. Training without sufficient access remains theoretical. Access and training without professional mentorship may produce technical ability, but they cannot reliably produce production judgment.
When I entered computer graphics in the mid to late 1990s, access was the weakest pillar. A serious workstation, specialist software, storage, video equipment, and the infrastructure required to operate them could cost more than several junior salaries. Most aspiring artists could not recreate a professional VFX environment at home. The systems existed inside studios, so anyone who wanted meaningful experience first needed access to one.
That economic barrier forced studios to participate in creating their own workforce.
They hired people who were not fully trained because fully trained people were rare. They looked for visual ability, curiosity, discipline, patience, and the capacity to learn. Then they gave those people access to equipment, software, experienced colleagues, and actual production.
This was not necessarily generosity. Studios trained people because they had little practical alternative. If they needed artists, they had to help produce them.
The old system should not be romanticised. Training was often informal, inconsistent, and dependent on whether a senior artist had the time, patience, and temperament to explain anything. Some juniors found generous mentors. Others were left alone beside expensive machines they barely understood.
It was not an educational ideal. But it placed beginners inside the environment where professional judgment could develop.
The studio provided access because the tools were there. Training emerged through manuals, internal instruction, experimentation, observation, and correction. Professional mentorship happened because beginners worked near experienced artists and could see how decisions were made under production pressure.
You learned that VFX was not a collection of isolated techniques. It was a chain of consequences.
A tracking decision affected layout. A weak asset created problems in lighting. A careless render choice transferred unnecessary work into compositing. Poor communication between departments could waste days. A small error introduced early enough could spread through an entire production before anyone recognised where it began.
Documentation could explain a function. Training could teach a workflow. Professional mentorship showed why a technically correct solution might still be visually wrong, economically impossible, or dangerous to the people working downstream.
That is the knowledge mentorship transfers. It teaches diagnosis, proportion, consequence, and responsibility. It reveals why the obvious solution may be the wrong one, why an apparently minor compromise may become expensive later, and why solving one department’s problem by transferring it to another is not really a solution.
Today, the conditions around the first two pillars have changed beyond recognition.
Artists can own computers capable of serious professional work. They can install sophisticated software, practise at home, build personal projects, test workflows, and spend thousands of hours developing technical confidence without waiting for a studio to open its doors.
Training has expanded just as dramatically. Schools, specialist teachers, online courses, tutorials, documentation, communities, and shared knowledge have opened VFX to people who once would have been excluded by geography, money, or distance from a major production centre.
This is genuine progress.
Good teachers provide structure, fundamentals, criticism, artistic direction, and technical understanding. Tutorials can explain methods with a clarity that was almost impossible to find in the 1990s. Personal work can reveal taste, curiosity, persistence, and initiative. Schools can create demanding assignments, collaborative workflows, deadlines, and production exercises that prepare students for professional work.
Artists can now arrive far better prepared than many earlier generations could have been before their first studio job.
The failure lies in what parts of the industry concluded from that progress.
Because access to computers and software moved increasingly into the home, studios began acting as though access to professional production was no longer part of their responsibility. Because training became available through schools, teachers, courses, communities, and self directed learning, they began acting as though professional development could be completed before employment. Because applicants could build polished reels independently, preparation was increasingly mistaken for experience.
The first two pillars became stronger, and the industry used their success as an excuse to neglect the third.
That is the structural mistake.
A personal project can be difficult, ambitious, and technically sophisticated. But the artist still controls most of the conditions. The camera can be changed. The asset can be rebuilt. The timing can be adjusted. The weakness can be moved outside the frame.
Production replaces much of that control with dependency.
The edit changes, but the deadline does not. The plate is incomplete. The camera data is wrong. The asset arrives late. Notes contradict one another. A beautiful solution creates an impossible render cost across eighty shots. A quick repair in one department transfers days of work into another.
The artist must understand what can change, what cannot change, who depends on the result, and what the proposed solution will cost elsewhere.
That understanding does not arrive automatically with better equipment or more training. It develops through responsibility, correction, and guidance inside actual production.
Professional mentorship makes invisible knowledge visible.
Much of that knowledge cannot be reduced to a tutorial because it is conditional. It depends on the shot, the sequence, the pipeline, the schedule, the client, and the people involved. The same technical problem may demand a different answer depending on what came before it and what must follow.
This is not a criticism of VFX education. The best teachers prepare artists for production, simulate many of its pressures, expose students to professional methods, and prevent years of confused experimentation. They make artists more capable of learning once they enter a studio.
But education cannot replace sustained responsibility inside commercial production at the scale only the industry itself can provide.
Education prepares artists to enter production. It cannot release production from the responsibility of developing them further.
Yet junior artists are increasingly expected to arrive already formed. They acquire their own equipment, obtain software, pay for courses, study in their own time, create personal work, rebuild their reels, and learn whatever new tool has become fashionable before anyone will consider hiring them.
The training requirement did not disappear. A growing share of its cost, time, and risk moved onto the individual.
This transfer is then disguised through the language of character. Unpaid training becomes passion. Financial access becomes talent. Available time becomes ambition. A studio reduces an expense and then treats the artist who absorbed it as unusually dedicated.
Even after all that preparation, the third pillar remains missing.
A junior artist can arrive with excellent equipment, substantial training, and an impressive reel. That artist may be exceptionally well prepared to enter production.
They are not already experienced in production.
The industry has mistaken unprecedented preparation for completed professional formation.
Its economic structure makes that mistake attractive. Studios bid aggressively, operate on narrow margins, expand around projects, contract after delivery, and move work according to tax incentives, client demands, and available capacity. Crews are assembled to solve immediate production problems rather than developed to build future capability.
Within that system, mentorship appears inefficient.
A senior artist is measured by output. Time spent explaining decisions, reviewing junior work, allowing controlled mistakes, or gradually increasing responsibility can look less productive than completing the task directly.
Mentorship is therefore expected to happen invisibly, without affecting delivery.
That is not a development system. It is an additional burden placed on experienced artists who are already carrying the production.
A rushed introduction to a pipeline is not mentorship. Documentation is not mentorship. Occasional review notes are not mentorship. Professional formation requires continuity, explanation, observation, correction, and enough time for responsibility to grow.
Without those conditions, juniors may learn how to satisfy individual notes without understanding the decisions behind them. They may become faster without becoming more perceptive. They may complete tasks without learning how to recognise when the task itself is based on a faulty assumption.
This is not true of every studio. Some companies continue to invest seriously in junior artists and understand that professional development requires time, continuity, and experienced supervision. But across an industry organised increasingly around immediate delivery, mentorship is too often treated as optional work that should happen after everything measurable has already been completed.
Senior artists conceal the damage. They repair weak shots, explain missing fundamentals, resolve failures between departments, protect schedules, and absorb errors created by people who were never given enough professional guidance to develop.
Their competence makes the system appear healthier than it is.
When experienced artists rescue a production, management may interpret delivery as proof that the structure worked. In reality, the production survived because experienced people compensated for structural weakness. The client sees a completed sequence, the industry sees no emergency, and the cost is carried invisibly by the people preventing collapse.
AI may intensify this problem, but it did not create it. Used well, automation could remove repetitive labour and create more time for feedback, observation, experimentation, and gradually more demanding work. The danger begins when studios keep the efficiency gain but discard the developmental function that once surrounded the task.
Roto, paint, cleanup, tracking, preparation, and simpler compositing work were not only production activities. They were also routes into real pipelines. They gave junior artists contact with plates, motion, lenses, continuity, versioning, deadlines, and the discipline of delivering work that other people depended on.
There is no reason to preserve inefficient work for sentimental reasons. But when those tasks disappear, their developmental role must be replaced deliberately. Otherwise, automation removes another place where professional mentorship could begin while leaving the demand for experienced artists untouched.
AI is only one pressure among many. Short contracts, aggressive bidding, outsourcing, unstable crews, compressed schedules, and the expectation that every applicant should arrive fully formed had already weakened the system long before generative tools entered the discussion.
The central failure is simpler.
Access makes sustained practice possible. Training makes that practice informed. Professional mentorship turns preparation into judgment under real conditions.
The three pillars are not alternatives.
For a while, VFX can continue living from experience created under earlier conditions. Current senior artists can supervise, correct, improvise, and rescue. Their knowledge can conceal the fact that the structure beneath them is no longer producing enough people capable of replacing them.
But experience cannot be purchased the moment it becomes necessary. It takes years to create. It grows through practice, training, mentorship, failure, responsibility, and repeated exposure to the consequences of real decisions.
VFX is currently consuming stored competence. It is spending knowledge created by an environment it has largely abandoned while describing that abandonment as efficiency.
Eventually, experienced artists will retire, leave, burn out, or become too expensive for the budgets that still depend on them. The industry will then try to recruit its way out of a problem it spent decades creating.
It will search for senior artists who were never given the chance to become juniors, supervisors who were never given responsibility, and production judgment that no school, tutorial platform, software licence, or automated system could provide alone.
The first two pillars cannot rebuild the third after the people capable of providing it are gone.
By the time the industry understands what it has lost, the missing expertise will not be waiting to be hired.
It will never have been created.