DAR ES SALAAM: PERMIT me to begin with a confession, because every respectable Friday read should start with somebody admitting something mildly embarrassing before advising the nation. I belong to the generation that believes most photographic problems can be solved by searching a camera bag.
Beneath a dead battery, three mysterious cables, a receipt from a shop that closed in 1998 and a lens cap last seen during the Mkapa administration, there must be an answer. There may also be a boiled sweet of uncertain age, but eating it would require approval from a doctor and the National Archives.
For much of my professional life, a missing photograph meant travelling to find it, befriending an archivist or waiting until Monday. Then Artificial Intelligence arrived without a camera, tripod, generator, driver, reflector holder, microphone assistant or lunch allowance. Most suspiciously, it announced that it could turn a few paragraphs into moving pictures.
Nobody even had to shout, “Quiet on set!” while a neighbour’s rooster provided commentary. Naturally, the doomsayers arrived looking like people who had just seen a goat occupy the chairperson’s seat at a board meeting.
They declared AI fake, dangerous, lazy, foreign and demonic. Some announced this through WhatsApp voice notes beginning with coughing and ending with “forwarded as received”. Others uploaded warnings to TikTok with dramatic music and a beauty filter while doing community service.
There is something comic about using an algorithmic platform to announce that new technology must be rejected because it is unnatural. These are often the same people who once said digital cameras were toys, mobile money was unsafe and social media would disappear by Christmas.
ALSO READ: The Spain blueprint: How Taifa Stars can shock Africa in AFCON 2027
Today, they forward “Good Morning” flowers before sunrise and panic when TikTok takes four seconds to load. That is why AI-generated filmmaking matters, especially in Tanzania, where the greatest obstacle to storytelling has rarely been imagination.
Our problem has usually been money, equipment, access and preservation. Suppose you want to produce a serious documentary about the TanzaniaZambia Railway Authority, TAZARA. You have researched its politics, engineering, sacrifice and Pan-African ambition. You have interviewed retired drivers, engineers, labourers and villagers who watched the railway cross the landscape. Splendid.
Now provide thousands of workers dressed accurately for the 1970s, vintage locomotives, period machinery, authentic stations and enough money to feed a crew for three months without causing a constitutional crisis in the accounts department. This is the department where purchasing printer ink may require three quotations, four signatures and diplomacy normally reserved for border negotiations.
At that point, most producers return to black-and-white photographs, slow zooms, dramatic music and a deep voice saying, “It was a time of great change.” In documentary language, this often means, “We could not afford the reenactment.” There is no shame in that. Tanzania’s history has frequently survived because somebody’s aunt kept an old biscuit tin beneath the bed. Inside may lie the only surviving photograph of an important event, resting between buttons, receipts and a key that opens nothing known to mankind.
Artificial Intelligence offers another possibility. Take the surviving photographs. Add maps, engineering plans, newspaper reports, oral testimony and proper research. Then use AI to reconstruct missing moments: surveyors marking the route, workers laying sleepers, engineers discussing bridges and villagers watching a locomotive arrive for the first time. Not to invent history.
Not to replace evidence. Not to allow imagination to put on a safari suit and claim it attended the ceremony. The purpose should be to reconstruct carefully, label honestly and illustrate responsibly.
AI filmmaking has expanded because it reduces the cost of dreaming. Until recently, cinematic storytelling belonged mainly to people with serious budgets, expensive equipment and walkie-talkies used mostly to ask, “Where is the driver?” Today, a graduate in Mbeya with research skills and a capable computer can create scenes that once required a television station.
Photography did not eliminate painting, television did not eliminate radio, and digital cameras did not eliminate photographers.
Smartphones merely produced billions of unnecessary pictures of lunch and car dashboards. What technology usually destroys first is monopoly: the belief that only a small, wealthy group has permission to create. The greater opportunity, however, is preservation. Tanzania has no shortage of history. We have a shortage of organised documentation.
Consider Bongo Flava: three decades and two years of music, fashion, language and social change, yet surprisingly little has been systematically archived. The same applies to architecture, sport, broadcasting and everyday life. How many young Tanzanians can describe Beda Amuli’s work or have seen the founding years of Clouds FM beyond a few fading photographs?
ALSO READ: How football made Karia
These stories deserve documentaries. And or the first time, producing them may not require Hollywood. It may require a serious researcher, a careful writer, surviving photographs, eyewitness testimony and a laptop that does not sound like a departing aircraft whenever one opens four tabs.
But let us not become drunk on possibility. AI has one dangerous talent: it lies beautifully. Give it a careless prompt, and it will produce a convincing historical scene with the wrong uniforms, buildings, flags and people who never existed. It can place a leader beside machinery built twenty years later with the confidence of a witness who was not born but still insists, “I was there.”
That is why journalists, historians and filmmakers remain essential. A machine can generate a railway worker. It cannot interview the real worker whose back still hurts during the rains. It can recreate an old station, but it cannot understand what that station meant to a village.
It can imitate a speech, but cannot determine whether it was ever delivered. Research, judgment, context and ethics remain human duties. Every AI reconstruction must be clearly labelled. Audiences must know whether they are seeing original footage, dramatic recreation or generated imagery.
And, mind you, the label should not appear briefly in lettering smaller than bank-loan terms and conditions. So, to the doomsayers: nobody is asking you to worship AI. Suspicion is healthy. Journalism itself is organised suspicion. But jeering at technology without learning how it works is not wisdom. It is fear wearing reading glasses and introducing itself as experience.
One cannot dismiss the future as nonsense while spending six hours daily on WhatsApp and TikTok, platforms powered by AI algorithms deciding what one sees and which dancing stranger interrupts the afternoon. Artificial Intelligence is not waiting for us to believe in it. It is already here. The question is whether Tanzanians will use it to tell Tanzanian stories, preserve Tanzanian memory and create opportunities.
Or sit beneath the mango tree complaining until somebody abroad produces the TAZARA documentary, gets our history wrong, adds a lion to every scene for “African atmosphere,” and wins an international award for authenticity. I remain old-fashioned enough to love original photographs, dusty archives and family albums. Nothing replaces evidence.
But AI offers historians and filmmakers a new tool: not the right to rewrite history, but the ability to help people see what history may have looked like. Used dishonestly, it will confuse.
Used lazily, it will produce rubbish with cinematic lighting. Used responsibly, it may make the silent archive speak. And should the next great TAZARA documentary begin not with a camera crew boarding the train, but with a historian opening an archive and an AI model waiting beside the keyboard, filmmaking will not have died. It will simply have changed trains.