The Screen Got Brighter. The Shadows Got Bigger

The most revealing thing Vivek Kamath did in this conversation was sort his own answers. Twice he prefaced a response by telling me which level of the pyramid he was answering at. The top of the pyramid first, he said, and then I will address the entry level. The structure of his thinking is itself the system this episode examines: an industry that organizes its talent hierarchically, treats its hierarchy as natural, and gives most of its analytical care, legal protection, and reskilling promises to the top.

The system is the contemporary Indian entertainment economy as it is being reshaped by streaming, data-led production, and AI. Vivek is the founder and director of Matrix Entertainment, India’s largest celebrity management agency, which represents Alia Bhatt, Priyanka Chopra Jonas, Ram Charan, Vicky Kaushal, and a long roster of the top tier of Indian film. He sees the industry from the top floor, and from there the changes look like a story of adaptation. From elsewhere in the building, the same changes look different. The question is what gets hidden by the view from above when that view is the one being asked to describe the whole house.

The default participant the industry’s protections were built around is a star or an established lead. Their image rights are contractually negotiated; their AI likeness is legally defensible; their representation, by an agency like Matrix, ensures that any use of their face or voice comes with paperwork and compensation. Vivek illustrates this with Natalie Portman: a studio that scanned her for Black Swan can use that scan for Black Swan and a sequel with renegotiation, but cannot place a copy of her in the next Avengers film. The protection works because Portman is at a level of the industry where lawyers, contracts, and leverage exist.

The default does not describe the daily-wage extra, the junior writer on a first credit, or the makeup assistant trying to break in. None of them have a lawyer reviewing their image-rights clauses. Most of them do not have image-rights clauses at all.

Vivek’s pyramid framing gives the audit two distinct exclusions to follow, separated by who can survive automation through “adaptability” and who cannot.

The first is the entry-level performer whose career used to begin in the spaces automation is now consuming. Crowd actors, background performers, extras. Vivek concedes this directly and immediately minimizes it: in percentage terms, he says, it is not a significant amount of talent, and these are daily wagers who likely have another job. Both claims are true. Neither addresses the question of what happens to a profession when its entry-level disappears. A film industry that no longer needs crowd scenes does not just shed crowd actors. It shuts a door that has historically been one of the few open doors for working-class aspirants without an academy credential, a manager, or a network.

The second is the technician whose multi-skilling Vivek treats as the answer to displacement. He describes a future in which a set technician who worked on one film a year retrains and does ten to fifteen YouTube productions. The hours may add up. The income might not. The benefits, the stability, and the union protections of film-set work do not transfer to creator-economy gig labor. Multi-skilling, named as a virtue, is also a description of an industry that has stopped providing stable single-role employment.

A third layer involves Vivek’s discovery argument. Everyone can find a platform, he says. Discoverability requires social media presence, consistent output, and marketing investment, which presupposes a creator with savings, time outside paid work, and the social capital to navigate the platforms. The newcomer without these is invisible to the discovery pipeline. The same selection logic operates in influencer casting, where follower count substitutes for training and structurally erases the trained performer without a social media presence, the woman whose family limits her public visibility, the person whose talent has never been monetized into followers.

The entertainment system in this conversation optimizes for scale, engagement, and measurability. The numbers it watches are stream counts, follower counts, opening-weekend grosses, completion rates, ROI on individual films. These are the metrics Vivek’s industry can read clearly, plan around, and bet on. They are also the metrics that reward the top of the pyramid, where scale concentrates, where engagement is measurable, where the data is clean.

What this measurement model under-protects is craft, access, and consent. Craft, the on-set knowledge accumulated by location scouts, practical-effects artisans, and junior writers learning the room, does not have a metric and does not show up in the dashboards that drive commissioning. Access, the question of who gets through the door of the industry at all, is influenced by discovery and casting signals the platforms favor, which means the door narrows in the direction the data points. Consent, particularly for AI likeness use, is treated as a problem for those with legal resources and a non-problem for those without. The pattern is not indifference. The industry has built a sophisticated system for what it can measure and a thin one for what it cannot, and the people inside that system describe the gaps as adaptability problems rather than design choices.

Three redesigns follow, all at the bottom and middle of the pyramid where the existing apparatus is weakest.

The first concerns automatic, standardized residual rights for any performer scanned for AI likeness use, regardless of star tier. Vivek’s Black Swan example shows the rights framework exists; what is missing is its universality. Background actors scanned to populate digital crowd scenes, junior performers whose likeness is used to train generative models, even the daily wagers Vivek describes as “not significant in percentage terms,” should have non-waivable rights to know how their image is used and for how long, with default compensation triggered by reuse. A studio might argue this raises costs. It would, marginally. Studios already accept those costs for stars. The redesign is to extend the apparatus, not invent it.

Furthermore, reskilling should be paid for by the entity adopting the automation, not by the worker being displaced. Vivek treats reskilling as a worker disposition: technicians will reskill, he says, because they are multi-skilled. He leaves unsaid who pays for the courses, the equipment, the time off-set during which the worker is not earning. A redesign worth its name treats reskilling as a cost of automation adoption rather than a private inconvenience. A studio or platform that displaces a category of worker should fund the transition out, on the same financial line as the technology that produced the displacement.

Finally, mandatory commissioning slots for new voices on streaming platforms. AI-driven commissioning is risk-averse by design: it reads past performance and bets on whatever the data points toward, which in practice means more of what already worked. The result selects against newness. A redesign requires that a defined percentage of every platform’s commissioning slate, say fifteen percent, goes to first-time directors, writers, or on-screen leads. The objection will be that quotas distort the market. The current data-driven commissioning is also a distortion, and its distortion is invisible because the algorithm performing it is not labeled as one.

The most useful thing I noticed was a habit of structure. Vivek would acknowledge a problem at the entry level, give it a sentence, and then walk back up to the top of the pyramid to explain why the larger system was fine. The reassurance always arrived faster than the acknowledgment had time to settle.

I am not sure he was doing it on purpose, and that is part of what made it striking. The optimism might not be strategic; it could be genuine, and it is the voice of someone who has spent his career representing the people for whom the system continues to work well. An optimism that nonetheless genuine forecloses certain questions before they can be asked. If the arts will always find their way to the light, then we do not have to look very hard at who is being asked to do that finding, with what resources.

By the end of the interview I had started counting the word “adaptability.” Each time it appeared, something difficult was being closed up rather than examined. It described a virtue. It also absolved an industry.