Innovation vs job losses in local AI film 'Mountain of Graves' - TimesLIVE

TEXT START: South Africa’s film industry has been making great strides in the tech space.

The Dissection

The article presents AI filmmaking as democratization: lower production costs, smaller teams, faster iteration, and greater access for African creators. It accurately describes a hybrid workflow in which humans still direct, edit, curate, and supply some creative judgment.

But its deeper function is normalization. It reframes labor substitution as entrepreneurial opportunity and treats the survival of a few creative decision-makers as evidence that the broader workforce remains secure. The workflow is not merely a new toolset. It is a compression engine for production: physical sets, cameras, actors, voice actors, sound teams, and parts of post-production become software functions or optional inputs.

The “micro-studio” is not a rescue of the old film economy. It is a smaller ownership layer operating above a much thinner labor market.

The Core Fallacy

The central error is confusing more output and more creator opportunities with more economically necessary human employment.

AI can make feature-scale content cheaper while requiring fewer people to make each feature. That may create new roles for directors, prompt designers, editors, IP owners, audience builders, and technical supervisors, but those roles do not scale in proportion to the displaced workforce. The system can produce more films while purchasing less human labor overall.

The claim that AI may become a “net job creator” relies on gross job creation while ignoring substitution, wage compression, and competitive selection. A micro-studio may employ a small creative core, but its cost advantage is precisely that it avoids the large employment pipeline of conventional filmmaking. The article mistakes a new production architecture for an expansion of mass productive participation.

Under the Discontinuity Thesis, this is the early form of P1: cognitive and creative automation becomes cheaper and sufficiently capable. P2 follows as institutions fail to preserve human-only production domains against cheaper AI-native competitors. P3 arrives when most workers can no longer access economically necessary labor, even though content production itself expands.

Hidden Assumptions

• That human involvement means human labor will remain economically indispensable. It does not. A person can remain “in the loop” while the number of people required in the loop collapses.

• That lower barriers to entry benefit workers more than they benefit owners and platforms. Lower barriers usually create an oversupplied field, intensifying competition and reducing the price of creative labor.

• That audiences and distribution will expand fast enough to absorb the flood of inexpensive AI content. Attention is scarce even when production is not.

• That human authorship will remain protected by market preference. Some premium audiences may demand it, but that creates a niche moat, not a mass employment guarantee.

• That newly created roles will be broadly accessible. The likely beneficiaries are owners of IP, distribution, compute access, audience networks, and high-leverage technical or creative judgment—not the displaced production workforce as a whole.

• That “net job creation” is the relevant metric. The relevant question is whether ordinary workers retain bargaining power and access to necessary income. The article does not establish either.

• That Africa’s disadvantage is mainly a lack of production capital. Capital scarcity is part of the problem, but ownership, distribution, infrastructure, energy, and market power determine who captures the gains after AI lowers production costs.

• That regulation can contain the disruption without confronting the ownership structure. Regulation may slow deployment or protect likeness and copyright. It cannot restore the old labor-to-consumption circuit once cheaper AI production becomes competitive.

Social Function

Primary classification: transition management and ideological anesthetic, containing a partial truth.

The partial truth is real: AI filmmaking can let small African teams create work previously blocked by cost, equipment, or access constraints. It can create temporary niches for directors, technical curators, IP owners, editors, audience operators, and other high-leverage specialists.

The anesthetic lies in presenting these niches as evidence that the employment system is being renewed. The article converts a reduction in required labor into a story about empowerment. It celebrates the fact that a small team can now do what a large crew once did, while treating the missing crew as an undefined “disruption.” That is not neutral language. It is the vocabulary of managed displacement.

The micro-studio model also shifts risk downward. Workers are encouraged to become founders, creators, or audience-builders, while ownership of the resulting IP and distribution channels becomes the decisive source of power. Most entrants will not become Sovereigns. They will compete for unstable servitor roles in an oversupplied attention market.

The Verdict

The article correctly identifies AI filmmaking as a powerful cost and access breakthrough. It fails at the systemic level. AI may produce more African films, more micro-studios, and a handful of new high-value creative positions while destroying the labor density that sustained conventional film employment.

This is not evidence that AI preserves the film industry’s job base. It is an early demonstration of how the old model dies: production survives, output rises, ownership concentrates, and the workforce becomes expendable. The winners will be those who own the IP, audience, distribution, compute, energy, or logistics. Everyone else is competing for a shrinking number of human-necessary positions inside an expanding machine.