It’s your first day of a new job. HR ask you to do something you never expected: sit in a circular booth, repeat a list of words from an autocue, and be filmed from all sides.
Your digital image and voice will be captured and then manipulated to place you anywhere and to have you say anything that an AI model deems appropriate. Your image and voice will be used in corporate communication videos while you work for this company. You’ll never need to sit in front of a camera again. But you might want to consult a lawyer.
It’s a brave new world, one beset by technical and ethical challenges, but one that also presents new and exciting opportunities.
There are challenges and issues around the use of AI, but every day, film companies the world over are finding fascinating new ways to streamline the production process and cut costs even as they reach technical heights unimaginable before the arrival of machine learning technology.
Video production companies are no strangers to new technologies. Over the last twenty years, we’ve seen almost unimaginable advances in everything from cameras and connectivity, to developments in post-production like automated transcription allowing ‘paper edits’ and automated voice generation. However, AI is something more fundamental that has the potential to transform not just how content is produced, but also what we choose to create, whether that is informed by costs and efficiencies or an algorithm determining what is most likely to achieve our goals.
What is AI?
Artificial intelligence is any programme or device which can mimic functions and processes typically associated with humans, like “perceiving, reasoning, learning, interacting with the environment, problem-solving, and even exercising creativity”. As is typical of any technological gold-rush, wild claims are made and while the potential applications are astonishing, it’s important to distinguish between what’s genuinely revolutionary and what’s just a more efficient way of working.
As it stands, there are already very effective workflow tools, using certain elements of AI, for instance, voice recognition in the generation of video transcripts. The AI model has been taught to relate words to sounds, analysing sentence structure and context by trawling millions of pieces of content. This clearly obviates the need for someone to do a time-coded text transcription of, for instance, interview content. You just push a button and it’s there – attached to the rushes, searchable, updating along with the edit and just another button push away from generating captions. It’s nothing new, just a quicker way of doing something.
In addition, there are generative tools, like those we explore below. These are actually using machine learning to create. That is new.
So how do they work? Let’s take an obvious example: ChatGPT. OpenAI’s flagship product was originally raised on a steady diet of content – including text, images, video, audio, and pure raw data – from across the internet. An algorithm built into it allowed it to observe patterns and associations in what it was shown. So while it may, at first glance, simply be scraping and aggregating existing data, the way in which it uses this data allows it to create something new, if not entirely unique. This goes far beyond writing undergraduates’ essays for them. As everything in our modern world is essentially bits of information, AI is capable of making decisions about how things should be created, everything from generating images, to structuring stories, editing film and, of course, determining what is deemed successful to further enhance its efficacy.
The holy grail is, of course, a machine that is genuinely creative. As the development of AI continues at a rapid pace, this third and final type of model, an autonomous machine requiring no human input, is within reach. Will we like what it’s thinking and what it chooses to believe? How we evaluate its thoughts and output, and if we see it as being truly sentient, really depends on our understanding of what makes us unique. Some people like Avengers:Endgame, some prefer the subtle nuance of The Bicycle Thief. Will AI be the same and, if so, which bot will we prefer? As content becomes increasingly homogenised, and feedback loops ever more sophisticated, it’s an important question for the creative process. And we’re sure the machines will have an opinion about that.
Generative AI in Content Creation
The Issues
Taking time and costs into account, AI-generated content offers greater efficiency in the production process. However, there are many issues to be addressed, not just in terms of audience expectations and authenticity, but also legal and ethical standards covering everything from copyright and liability to online political messaging.
Some AI-generated content can feel sterile, bland, or generic—lacking nuance and emotional depth. Take, for instance, deepfake technology, where faces and voices are manipulated to create hyper-realistic, unsolicited representations of people. Setting aside the legal implications, there’s also the issue of how AI-made footage looks. You may have heard of the uncanny valley – the feeling you get when something seems human or human-made, but just doesn’t feel quite right. AI offers some prime latter-day examples with de-ageing technology used in major studio movies.
When Martin Scorsese’s not-for-the-short-of-attention The Irishman was released in 2019, one of the most-commented on aspects (aside from the runtime) was the de-ageing technology used to transport Robert de Niro back in time to the 1940s.
The image was accurate, most agreed, but the motion wasn’t quite right – “an old person doesn’t move like a young person,” noted visual effects artist Joe Pavlo. “They don’t walk the same. Their mannerisms are not the same.” The same problem has cropped up with de-ageing treatments given to Mark Hamill, Jeff Bridges, Harrison Ford, and Tom Hanks.
Aside from aesthetic concerns, AI models can inherit and amplify the biases present in the data used to train them. Whether it’s racism, sexism, or other forms of discrimination, AI-generated content can perpetuate these biases if left unchecked. This can manifest in everything from casting suggestions for digital characters to the way AI generated scripts are written.
Moreover, factual errors are a significant issue, as AI models can confidently produce incorrect information if not thoroughly monitored. Lastly, when AI is used without clear creative direction, the content may fail to capture the original intent, values, and unique style of the filmmaker or creator, making it appear directionless and inauthentic – or even just weird.
Some creators have embraced the unreal and leaned into the otherworldliness of AI imagery. Dutch artist Jeroen van der Most, working in what he calls “Dystopian Weirdcore Surrealism”, has used AI tools to create unsettlingly hysterical videos featuring strangely realistic eggs, people made out of pizza, and fantastic creatures with lifelike movements. Creator Supercomposite took the internet by unnerving storm with Loab, a bizarre AI-made woman with a scarred, bleeding face.
Others have coupled the technology with truly inspiring creativity to create a film we’d pay to see. Pirates of the Caribbean meets the 1950s in glorious technicolor? We’re in.
On the other hand, AI tools can be a force for change. When Nicolás Maduro clamped down on Venezuelan media in the summer of 2024 following a highly controversial presidential election, journalists in a country where “being on camera is no longer sensible” turned to AI to reach audiences. Operation Retweet, an initiative of Colombia-based journalism platform Connectas, uses AI-generated newsreaders to combat disinformation spread by the Maduro regime. “Before we go on – in case you haven’t noticed – we want to let you know that we aren’t real,” says newsreader La Chama (Bestie) as she’s deployed to allow journalists to protect themselves from identification and arrest.
Why It Matters
AI should serve as a support to enhance human creativity, rather than a replacement. For companies and creatives, the risk of losing control over the direction and message of their content is significant if AI is used indiscriminately.
The challenge organisations face is to harness the potential of these new technologies whilst ensuring that AI-generated content reflects their values, ethics, and vision, to strike the right balance between efficiency and creativity. In this way, AI can be a tool that enhances productivity, without sacrificing authenticity and originality.
They also need to think about the source of material being used to develop this content, not just its legitimacy, but also its provenance and legal issues arising from copyright and an increasingly complex legislative framework. The European Union is currently at loggerheads with some of the titans of Silicon Valley over its Digital Single Market Directive which seeks to protect copyright holders whilst promoting innovation. Some would argue that we’ve seen the decimation of the news gathering industry as its profits have been eroded by aggregators, scraping and repurposing content and then taking the lion’s share of the revenue. The commensurate effects in terms of scrutiny, originality and proliferation have had very real societal consequences. Do the creative industries now face this same threat? Are we all at risk of becoming just a little bit binary?
Streamlining the Production Process
AI is already here and key in a wide range of tools, platforms, and services which can simplify the production process and cut costs. The power to create imagery, footage, and audio, to storyboard and brainstorm, and to edit material using tools like predictive storytelling to preemptively connect with audiences traditionally called for specialists and many, many hours of intensive, costly work.
AI hasn’t replaced humans in video production, but it has simplified and automated the process of generating the building blocks of a piece of work. That translates to more time saved, less money spent on each stage of production, and increasingly superior results as algorithms build on what’s been successful to inform future decision making, for us and them.
Below, take a closer look at some of the tools creators are using to make this happen. These are only a few of the available resources. The future promises more to come, but as you’ll see, there’s already a rich array of options to choose from to bring your project to completion while saving time and money.
ChatGPT
ChatGPT is an AI-driven language model that can generate dialogue, suggest plot points, or help flesh out character development. This makes it a great resource for scriptwriting, and its “Image Generator from Text” feature simplifies the once time-consuming and costly process of storyboarding.
Those who’ve used it know that the game-changing chatbot can come out with sometimes inaccurate, sometimes odd text. That just goes to show that it doesn’t replace human creativity. ChatGPT can be a valuable tool in brainstorming sessions, for creating initial drafts of scripts, and even proposing narrative structures based on specific genre requirements. It can also act as a powerful tool for speeding up the writing process, allowing writers to develop ideas quickly, enhancing productivity during pre-production.
Sora
Another upcoming product from OpenAI, Sora helps make film and video production a bit easier by handling some of the more laborious post-production tasks. Primarily a text-to-video generator, it can also analyse footage to help with things like cutting scenes, colour correction, and sound adjustments, so editors don’t have to do everything manually. It also tags and organises media files, which makes it simpler for filmmakers to locate the clip they need from what will often be hours of rushes. With tools for adding effects and creating captions, Sora can speed up the process of refining raw footage, allowing creators to focus more on the creative side of storytelling.
ElevenLabs
ElevenLabs specialises in AI-driven speech synthesis, offering filmmakers highly realistic voiceover and dubbing solutions. This tool can generate natural-sounding voices in various accents and languages, making it ideal for tasks like creating temporary voice tracks during post-production, prototyping animated characters’ voices, or even replacing or enhancing voiceovers for films in multiple languages. The platform also offers “Iconic Voices,” AI recreations of famous voices belonging to everyone from Burt Reynodlds to Marilyn Monroe.
By enabling quick voice production with minimal need for a human voice actor in the initial stages, ElevenLabs also allows filmmakers to test different vocal styles and tones without having to commit to a recording studio early in the process.
Adobe Firefly
Adobe Firefly generates images and visual content from text prompts, allowing filmmakers and designers to quickly create concept art, storyboards, and even visual elements like environments and textures without the need for traditional graphic design or manual editing. Particularly useful on the conceptual side of video production, Adobe Firefly can dramatically accelerate the production of creative assets in the pre-visualization and design stages, enabling filmmakers to visualise their ideas more efficiently and experiment with different aesthetic styles in real-time.
Luma Dream Machine
Luma Dream Machine is a text-to-video generator that uses AI to automate and enhance visual effects and CGI rendering, dramatically simplifying the process of producing high-quality, photorealistic environments and characters.
This tool leverages machine learning to analyse large amounts of visual data, reducing the need for complex manual VFX processes like 3D modelling and animation. Luma Dream Machine is especially useful in scenes requiring advanced visual effects, allowing creators to achieve cinematic-quality CGI at a fraction of the traditional time and expense. It enables directors to create dreamlike worlds or action-packed sequences with precision and speed.
Wonder Studio
This tool can automatically detect actors in a scene and replace them with CGI characters, handling tasks such as animation, lighting, and even compositing. Wonder Studio uses AI to simplify the process of integrating digital characters into live-action footage. What once required teams of animators and visual effects artists can now be done in a fraction of the time, allowing filmmakers to experiment with digital actors and characters without relying on green screens or complex tracking systems. Wonder Studio’s automation of these time-consuming tasks makes it a game-changer for filmmakers large- and small-scale alike, especially for productions that rely heavily on digital or animated characters.
The Takeaway
Tools like ChatGPT can help you quickly build out storylines and flesh out topical and narrative ideas, while Luma Dream Machine, Firefly, and Sora can help you see what those ideas might look like on screen. Wonder Studio can enliven your videos with eye-catching visuals, and you can top it all off with relatable, realistic voices from ElevenLabs.
AI offers outstanding resources for planning and ideation. It can cut costs and, used properly, can free up staff to focus on the creative process, sparing them from many of the more time-consuming tasks in video production, such as creating or trawling through stock footage and b-roll, but it’s not, yet at least, a substitute for human creative input.
If you’d told us twenty years ago that we’d be helping clients edit footage shot on a mobile phone and posting it on YouTube we would have been surprised. Furthermore, if you told us audiences would accept that, we’d be incredulous. As the way in which people consume and value content, more and more of which will be created by AI, changes, so their expectations and engagement will evolve. Imagine where we’ll be in 2044.
The production industry needs to embrace the opportunities that AI offers, whether that’s a creative bootstrap, working with algorithms in developing scripts and storyboards, using tools to streamline the post-production process, or going all out with fully AI generated visual content.
At NextShoot, we’re always looking for ways to reinvent and improve the video production process for our clients. If you want to know more, give us a call or get in touch online.
