I Myself Am The Sun
Gayton McKenzie's Blockbuster: Algo Scriptwriting, Social Cohesion & The Death Of Chance
by Roger Young
In Hyènes, Djibril Diop Mambéty's 1992 masterpiece, a wealthy woman returns to her impoverished village offering unimaginable wealth on one condition: kill Dramaan Drameh, the local shopkeeper who betrayed her as a girl. No one wants to kill the man, but the presumption that someone will sparks off a wave of money lending, conspicuous spending, and deal making, until many of the villagers owe others for goods, credit, and favours, making the outcome inevitable.
The National Film and Video Foundation (NFVF) recently put out a call for “Blockbuster” pitches. Minister of Sports, Arts & Culture Gayton McKenzie, before rearranging the agency, said, “We need Blockbusters. We need Mzansiwood.” Both are achievable.
With Chicco Twala's low-budget bubblegum films, the kykNET/Hartiwood box-office circuit, Netflix slates, and our previously booming international service industry, an organic Mzansiwood was already in full swing before the DTIC production rebate fell apart. The primary obstacles to Gayton's blockbuster are the DTIC's bureaucratic failures, and NFVF's own Sediba development system.
“Blockbusters are A-list talent, budget and high-level producers dependent. Development is probably the least important ingredient. The closest thing South Africa had to blockbusters was Leon Schuster's films.” — Pascal Schmitz, producer, Amariam Productions
The NFVF's total budget last year was R160 million; an A-list star costs $20 million per picture.
Alongside its usual production slates, the NFVF recently issued three specific calls: Animation, Once Upon a Family, and Blockbusters. Under the in-house Sediba Scriptwriting Training Programme, script development is taught through a rigid three-act framework with predetermined turning points, binary aggressor-and-victim character mechanics, and an obligatory Act III “message of hope.” The Once Upon a Family track (capped at R6 million) explicitly requires scripts to “avoid intense violence or mature content” and “focus on positive and uplifting themes,” while Becoming (capped at R4.5 million) limits eligibility strictly to coming-of-age trajectories for youth crews under thirty-five. In all three calls, any proposal that does not explicitly align with these predefined themes is effectively disqualified — if it makes it to draft stage at all. These operational guidelines reflect the Department of Sport, Arts and Culture's (DSAC) stated strategic policy, which defines the purpose of state-supported art not through formal or artistic criteria, but through measurable socioeconomic outcomes: “social cohesion, nation-building, job creation, and economic empowerment.” But the NFVF develops films that do not cause debate, are tightened against controversy, yet the very job of a state film agency should be to support work that risks causing controversy.
To sustain this funding model, the NFVF requires that its production investment be recouped first — in full — before deferred fees, deferred costs, or back-end participation are paid. Yet the NFVF's own annual box-office reports undermine the premise: local feature films funded under this system rarely recover their R4.5M–R10M budgets through domestic theatrical release or local broadcast sales. In 2021, fourteen local releases earned roughly 1% of a R325 million total box office, with 95% of that revenue concentrated in three comedies; in 2023, eight domestic films shared R5.2 million — again less than 1% of the national gross, a sum no single film's budget could recoup from theatrical alone. The NFVF continues to publish aggregated annual and mid-year reports but has stopped providing granular per-film revenue figures for the individual features and slates it underwrites, disclosing only summary trends and top performers. The state mandates specific narrative templates to guarantee that projects remain accessible, socially cohesive, and “commercially viable,” yet the resulting films routinely fail to recoup.
If Sediba films consistently, over decades, fail to recoup, why is the model still the standard?
When the NFVF was submitted direct questions regarding the micro-budget call, Sediba metrics, and commercial recoupment for this article, the agency requested a three-month extension to respond.
This evasiveness fits a pattern. Since taking office, McKenzie has dissolved the National Arts Council, reportedly blocked the NFVF's attendance at Cannes in 2025, cancelled Gabrielle Goliath's Venice Biennale pavilion over its inclusion of a Palestinian poet killed in Gaza — leaving South Africa's official pavilion empty, wasting the already-paid Arsenale rental and preparatory costs running into millions of rands, though Goliath's work was shown independently nearby — and the NFVF council has suffered resignations and instability with six CEOs in two and a half years, most recently suspending acting CEO Onke Dumeko in June 2026.
According to the DIFF programme, Isiphethu workshops this year will teach filmmakers to use AI to build pitch decks and funder-compliant business plans. The listed focus is not craft but alignment. If applicants feed NFVF criteria into generative tools, the LLM would accelerate narrative conformity. An LLM optimises whatever objective function it is given. Having no intention of its own, it produces structurally precise, terminology-compliant applications that precisely mirror what the agency has been trained to reward. The technology is not the problem. The problem is that the institution has created a system where conformity is the most rational strategy for survival. The result is a generation of filmmakers who have learned, through decades of Sediba workshops, that the path to support is not through artistic risk, but through institutional legibility.
And the applicant arrives on set with a map made by a tool that carries no experience of reality. An actor cancels, a thunderstorm arrives, a line of dialogue lands flat, and the person carrying the map hasn't examined the intentions behind these choices, and might struggle to adapt.
This compliance loop directly risks audience engagement. When an evaluation system requires every script to demonstrate an explicit character transformation and an unambiguous message of hope within ninety pages, narrative resolution must be fully spelled out on the page before production even begins. The audience becomes merely a receiver, passive, the film unspools without drama.
This methodology stands in direct opposition not only to open, independent cinema, but also to surprising mainstream cinema — the kind that breaks out precisely because it refuses to resolve on predictable terms.
This is the problem of mandatory closure. By forcing all dramatic ambiguity to be resolved in the script to satisfy DSAC nation-building metrics, the institutional framework structurally excludes cinema where the audience does the work, essentially excising critical thinking, feeling, and the joy of cinema itself, the reason people pay to see films.
It can be argued that all stories fit a three-act structure, some end in hope, some in parable, some in ambiguity. But you can't prescribe the structure, and the hope, upfront. The writer, the director, their journey to reach that on their own is where cinema resides. The very nature of stories contains a three-act structure. Since the dawn of time it has been why stories are told, it's how they captivate us. But if that structure and its actual timing is prescribed to a specific ending, it eliminates the chance and surprise of storytelling.
Streaming in South Africa promised to restore chance to storytelling. For generations, many Black South Africans saw themselves on screen only through stories shaped by institutions that did not belong to them. Opening Netflix and seeing local stories, local faces, and local languages is itself a vital cultural shift — representation creates confidence and confirms that a life previously ignored is now visible. Yet visibility and complexity are not the same thing.
Algorithmic curation and institutional funding both create representation through prediction, rewarding what is already recognizable: familiar genres, character arcs, and emotional outcomes. The mechanism is the same whether the gatekeeper is code or a compliance panel: proven patterns get funded; emerging ones do not. When complete legibility becomes the primary measure of success, culture is reduced to something designed to be immediately understood rather than something that challenges audiences to discover new ways of seeing. The risk is not that people finally see themselves on screen; the risk is that they only see themselves in forms that have already been approved — turning cinema into second-screen wallpaper.
A society cannot manufacture social cohesion by removing uncertainty from culture, because the act of encountering uncertainty is one of the ways societies learn empathy.
How does compliance to a form that eliminates chance, discovery and critical thinking promote social cohesion?
Netflix has built a global business on algorithmic predictability, and it works: Mandla Dube's Heart of the Hunter (2024) hit #1 globally on Netflix with over 11 million views across seventy-five countries in its debut week, while titles like Silverton Siege and iNumber Number: Jozi Gold also reached global Netflix audiences. Netflix can afford this because it has billions in capital, global distribution infrastructure, and data on 260 million subscribers. The NFVF has R160 million and no distribution arm.
When a public funding agency with no theatrical infrastructure and no streaming platform adopts “blockbuster bankability” as a funding criterion, it is not duplicating Netflix's success. The blockbuster strategy requires either volume tolerance or content freedom. The NFVF has neither. And it is doing so with public money that is statutorily meant to support work the market would not otherwise fund.
Similarly, the NFVF's prescriptive new Animation track ignores that independent studios are already achieving blockbuster-scale success entirely outside the state apparatus. The CGI animated feature David (2025), produced by Sunrise in Noordhoek, Cape Town with no NFVF underwriting, was released on 19 December and came in second at the North American box office behind only Avatar, opening to $22 million and grossing $87.5 million worldwide. Variety hailed it as “a spectacular showcase for the technical capabilities of South African animation house Sunrise Productions.” It did not pass through Sediba. Hope was not mandatory.
David is not the only recent South African animated feature to succeed without the NFVF. The 3D animated feature Headspace (2023), produced independently by Luma Animation and The Ergo Company, earned R2.6 million at the South African box office — a significant return for a local animated feature, and proof that independence works at domestic scale as well as global scale.
The same is true on television. Shaka iLembe (Bomb Shelter for Mzansi Magic, 2023) had the highest-rated drama premiere in MultiChoice history and drew over 7.5 million DStv viewers across its first season — and has since won twelve SAFTAs from seventeen nominations, the most ever awarded to a single series. It is now in its second season. Developed over six years with historians and the late King Goodwill Zwelithini, funded entirely by MultiChoice, and made with no NFVF development support.
The question remains: if we already have Netflix, independent animation studios, and private broadcasters hitting blockbuster numbers entirely outside the NFVF apparatus, what is the actual purpose of the NFVF?
The institutional focus on prescribing thematic content overlooks the underlying economic infrastructure that previously sustained high-volume domestic production. The foundational catalyst enabling our organic “Mzansiwood” was not prescriptive scriptwriting criteria, but financial predictability provided by the Department of Trade, Industry and Competition's (DTIC) Foreign and Local Film and Television Production Incentive. By offering reliable cash rebates of up to 35% on qualifying expenditures, the DTIC incentive provided the financial floor required for independent producers to secure private equity and international co-financing. Over the past three years, however, independent industry bodies (the IPO and IBFC) have documented a structural paralysis within the DTIC scheme, citing unpaid rebate claims exceeding R600 million, multi-year payment backlogs, and administrative opacity that have paralyzed local producers and triggered an exodus of productions: producer Nimrod Geva says a R160 million TV production with foreign financing lined up, including from the BBC, was declared dead after the incentive froze (https://groundup.org.za/article/lights-camera-but-very-little-action-why/); The Castaways reportedly considered a R90 million KwaZulu-Natal shoot, with about 80 jobs attached, before filming in Greece and Fiji; and Recipes for Love and Murder's third season is reportedly weighing Ireland or Scotland over a return to South Africa.
With the DTIC owing R600 million to filmmakers, where is the third act redemption?
The foundational lesson of the South Australian Film Corporation (SAFC), which triggered the Australian New Wave in the 1970s and helped build an industry that now attracts nearly USD 2 billion in annual production expenditure. It did this by providing reliable financial infrastructure (repayable operating loans) while maintaining strict arm's-length non-interference in narrative content.
A thriving film economy is built through stable financial infrastructure and artistic independence, whereas attempting to substitute broken financial incentives with rigid, outcome-driven script development formulas systematically undermines the industry it seeks to grow.
Meanwhile, the films that have carried South African cinema to global prestige share a formal trait: none of them would have survived Sediba's development grid. Oliver Hermanus's Skoonheid won the Queer Palm at Cannes. Jahmil X. T. Qubeka's Of Good Report was banned on opening night by the Film and Publication Board, then won seven SAFTAs and AMAA Best Feature. This year alone, Jason Jacobs and Devon Delmar's Variasies op 'n tema took the Rotterdam Tiger Award, Sandulela Asanda's Black Burns Fast secured Berlinale Generation 14plus, and Nico Scheepers' black-and-white Afrikaans horror Hen swept Silwerskermfees before winning the Narcisse at Neuchâtel — the first African film to take that top prize in twenty-five years.
Furthermore, independence can pay off on different levels. My own independent 2014 short film Keys Money Phone cost R46,000 to produce, and generated roughly R130,000 in direct distribution and screening receipts.
Our state agencies are actively destroying the very industry they are statutorily tasked to support. During the reporting of this article, two emerging filmmakers communicated — one by email, one by direct message — that they could not speak on the record for fear that criticism would cost them their NFVF development funding. Their names are withheld here to protect their eligibility. This is the true cost of outcome-driven compliance: a climate of self-censorship and fear.
Gayton McKenzie, the DSAC, and the NFVF do not trust filmmakers or audiences, and that lack of trust undermines their ability to achieve their own objectives.
In an interview for a 1983 documentary, the Senegalese director Ousmane Sembène said: “Why be a sunflower and turn toward the sun? I myself am the sun.”
This was in response to being asked whether his films were structured to be legible to European audiences.