Threat or opportunity? Coming to terms with AI
Are generative AI tools cost- and time-saving devices enabling producers to offer embattled channels more for less, or is this tech coming to cut the creatives out of the business entirely?
Wesley Block
If there’s one thing people fear more than change, it’s the unknown.
When it comes to artificial intelligence (AI) and the content business, we know change is coming because it’s already here and it’s happening at extreme speed. One minute you’re being shown how good AI is at dubbing your gameshow into foreign languages, the next you’re watching as Sora shows the potential to create and visualise a ‘new’ one after a few prompts.
Who knows what things might look like even 18 months from now, but it appears as if the industry is approaching a fork in the road.
The best-case scenario is AI is something of a saviour. The collapse of the ad market, declining subscriptions, rapid inflation, pandemic upheaval and more have left the traditional TV business on its knees. Channels cannot commission new content because people aren’t watching it and advertisers won’t pay for it. Producers cannot make their shows at the budgets being offered and stay alive. What if AI could strip away the banal, menial, time-consuming, costly bits of making television, allowing prodcos to turn out more content, quicker, for less money? How timely, thank God you’re here.
Meanwhile, the worst-case scenario is if AI keeps accelerating at this rate, why on earth would these budget-poor channels need to pay producers, writers, talent, creatives at all? If you can literally say to a computer, ‘I’d like a six-part murder mystery please; the detective is a female from Scandinavia who wears a lot of chunky knit,’ and it can produce it for you with the precise audience desires factored in from the algorithm, there’s a big chance they’ll choose to pay one prompt engineer as opposed to a whole writers room.
Wesley Block is a filmmaker turned co-founder and CEO of Kino AI, an AI tool that organises footage, speeding up the process for editors – very much part of that best-case scenario. “Development is the least logical place AI will and could make its impact,” he said during a panel at Realscreen Summit in New Orleans in February.
“Whether it can and what’s possible versus what will actually be the reality is a distinction often missed. We approach the conversation passively: there’s this high level of technology that will advance, there’s nothing we can do about it, market forces will drive it and audience tastes will adapt. As opposed to the other side, which is people are the industry. What do we want it to look like? What innovations do we want to see and drive? That’s how we should approach it. In the technical sense, you should assume anything is possible if you give it enough time and data.”
Megan McGrath and Irad Eyal’s Floor Is Lava
Irad Eyal, co-creator of Netflix gameshow format Floor is Lava, has more recently turned his hand to launching Quickture, an AI editing tool that helps analyse footage and find the story within it. Speaking on the same panel, he said: “I’ve been doing reality television for 25 years. What makes reality TV special is we all know we’re dealing with real people. We have to go out there and find characters, film the real story, and that is going to continue, with AI tools helping. I don’t think you’ll see a situation where documentary characters are replaced by computer-generated ones.”
Benjamin Field, an exec producer at Deep Fusion Films in the UK, who is also on UK producers association Pact’s working group trying to navigate the ethics and practicalities of how AI can and should be used, said: “Just because something is possible, does that mean we should automatically jump into it? Creative people are in the industry because they want to be creative. I don’t really see a drive for us to kill our own industry by purely making things because they’re financially viable. Creative people will endeavour to continue to be creative. AI can form part of that as another tool. Do we want to use those tools? Because they may harm our industry. We have responsibility to ourselves and a future generation of TV and film makers.
“We’ve been playing in this area for a couple of years and, as a creative producer and writer, there is an element of AI fatigue already. I use certain tools to make processes happen but actually I’ve now got to the point where I want independent creative control back because I feel distant from the project. I want control of my show because it’s my show.”
Realistic? Optimistic? Naïve? We’ll only know in time, but it’s the context of the current economic situation in the industry that potentially tips the balance towards foe rather than friend. Believing we’ll do what’s good and right versus the financial incentives forgets the fact it’s called showbusiness for a reason.
Eyal said: “If you do want a prediction of doom then where I see doom coming to the industry is in scripted studios, because now an individual person can make a movie or TV show that has Marvel- or Star Wars-quality visual effects and a cast of known or invented actors. In terms of scripted and big-budget programming, you will see a flattening where anybody can create that. But I maintain reality TV might be immune because we have to go and film real people.”
However, Field countered: “Yes, you can make stuff like that for TikTok and get a couple of thousand views but it’s a very different market from TV and film, where broadcasters and commissioners have a duty of care. There are processes in place to make sure high-quality content remains high-quality content. It’s not going to accidentally happen.
“Broadcasters and producers have the ability to work within set guidelines. The conversations we’re having with the BBC, ITV and Channel 4 are about us taking control of this narrative, ensuring we behave responsibly and these things don’t happen. Legislation is very slow; we need to be self-motivated.”
Oz Krakowski
Oz Krakowski, chief business development officer at Deepdub.ai, an AI dubbing company aimed at helping content owners and producers unlock monetisation opportunities for their content in new markets, is also mindful of the legal framework currently lagging behind AI advances, but believes it will eventually “catch up.”
“All these things we can do but shouldn’t are eventually going to line up and the industry will get to a point where it’s defined what you can do and what you can’t. There will always be outliers but it will stabilise and extend our work, not be a dystopian future where everything is doomed.”
Let’s use this as a jumping-off point for the best-case scenario and look at where AI can help producers and broadcasters allay the economic fears that have led to a commissioning stalemate. Block said the most useful and impactful tools are the ones everybody is on the same page about – translation, searching footage, dubbing.
“If I had a production company, the first thing I would do is lay out the entire workflow step by step and look at where we can save time,” he said. “Time-saving should be the focus because it translates to saving money – you can deliver sooner, rent equipment for less time and do more projects in a year. For the creatives and editors staying up late every night to get work done, it matters a ton to their quality of life. The entire conversation around implementing AI into a creative workplace should be around saving time in the workflow.
“Post is the most important thing to focus on. Unscripted producers capture everything and then, once they have all the footage, they craft a scene in a creative way. If you shot it over months and years, where were the golden moments? Computers can be extremely helpful with that.
“If you’re in an edit room thinking, ‘She’s great, get me everything we shot of her on this day,’ that’s a menial task somebody would have to spend a day doing. The computer can do it [in an instant]. Once it gets to the AI saying, ‘I think this is the best one,’ then it’s wrong. I’m uninterested in automated decision making. We are the storytellers, but it’s very good getting me the options.”
Dark Slope’s unscripted animated series Babble Bop
Raja Khanna, founder of Canada-based Dark Slope, which uses AI to create TV content including unscripted and animated series such as Peacock’s Babble Bop, said: “To help with research, structure ideas, organise reams of data… you’re wasting time if you’re not using it today. That’s a fact. It can save you hours and hours a month for your research and development team.”
Eyal added: “If you can turn around a series in one month rather than nine months, suddenly you have a very different workflow and opportunities to tell all kinds of different stories. We started from a position of trying to fix our own problems. If you’ve got hours and hours of casting tapes, it’s a gruelling job trying to find the nugget to send to a network. AI is very good at finding good-quality material in the stuff you have. If a network says, ‘Let’s focus on this bit about his grandmother,’ you know that would be back to the edit suite for three days. But something like Quickture could produce a sample of that much quicker.”
Is this, however, doing away with exactly the sort of entry-level jobs people use to access an industry that’s notoriously hard to get into, particularly for minorities, working-class kids, and people who live outside major production centres – the exact people we’re meant to be trying to involve more?
Field told a story of a quick-turnaround doc his company produced where ChatGPT was able to provide information on a topic they knew little about, answers to potential interview questions and generate mock contributors with different voices to present at the pitch within two days.
“Have we done somebody out of a job? In that specific case, I don’t think we did,” Field said. “It’s not as if we had time or resources to hire somebody to do that work and find six voiceover artists. In that case, it was the difference between being able to deliver or turning the project down.
“In the UK, commissioning levels are way, way down. Producers will have to make choices between closing because we can’t afford to keep going or finding a cheaper way of doing it. AI offers that. Some jobs will disappear, other areas of the industry will grow. People who weren’t in the industry, like software engineers, will move in. Traditional roles may disappear and that will be awful for people everywhere. As with every evolution and revolution, it’s inevitable.”
Krakowski added: “If you don’t use it to your advantage, someone else will use it to theirs.” His point was echoed by Khanna, who said: “If we don’t, somebody will.”
This rather paints the situation as an either/or. As long as the ‘goodies’ embrace this and use it for pure means, the ‘baddies’ won’t get their hands on it and ruin everything. The reality is it will be both. For every embattled production company using AI ethically to cut costs and survive an economic crisis, there will be someone else using it to cut creatives out of the process entirely.
Phil Gurin
Telling C21 about his experience with a sitcom-writing friend who prompted AI to produce a comedy pilot script, format veteran and Frapa co-chair Phil Gurin says the show it came back with was “not that bad. In fact I’ve worked on worse.”
Gurin also believes AI is capable of not only producing a copycat of, say, Who Wants to be a Millionaire?, but also making just enough changes so it complies with Frapa’s scoring system on IP theft and avoids legal action – potentially destroying the formats business overnight. Bearing in mind markets such as China were embroiled in a series of format copycat rows before any of this, it’s inevitable some businesses are going to be using that opportunity if it’s available to them.
Perhaps you’re a linear channel facing budgetary pressures, or a small player trying to compete against consolidated Disney/Warner Bros Discovery-type behemoths. Maybe you are one of those giants trying to rabidly slash overheads, or you’re a channel/streamer in a growing market in East Asia or Africa.
For all those groups, using a computer to spit out a show for you in a fraction of the time and cost it used to take is going to hinge on whether the audience will be willing to watch it. As cynical as this may sound, if the audience buys it, it’s a no-brainer that big chunks of the industry are going to do it, regardless of the ethics.
Meanwhile, there may be bigger fish to fry, especially when you consider how AI tools are being used to put words into politicians’ mouths in a year when more than two billion voters are set to go to the polls in 50 countries, including the US and India.