Relatively speaking, AI video creation is in its infancy still with perhaps the most famous example being Sora AI Video Generator, which is an OpenAI video generator that garnered a lot of media attention for some of the creepy imagery spit out from user prompts. Despite strange faces, extra limbs, or almost surrealistic interpretations of user text prompts, AI video creation is only going to improve. For all of us in the film industry, the response is a mixture of curiosity and terror. Could AI threaten our livelihoods and replace our jobs? The knee-jerk reaction is to reject new technology if it could be a threat, but the wiser approach is to embrace new technology as merely tools in the creation of human-directed art and marketing videos. Ultimately, people connect with other people. Sure, maybe one day an AI video generator could spit out a 30 second Taco Bell commercial with all AI-created “humans” and no camera being involved at all, but even then, the commercial has to be specifically designed to appeal to consumers — actual humans, not AI creations.
At present, the best use for AI video creation tools is to make small clips used in larger, human-created projects or for throw-away social media engagement videos. For instance, platforms like Pictory.AI can take a script or article and turn it into an AI-generated video complete with stock footage, AI-generated voiceover, and text. With how SEO favors videos over text, there’s something to be said for throwing up easy-to-create AI videos of a press release or something along those lines instead of just a wall of text. The results are decidedly mixed, though, with a lot of the AI-generated results needing some massaging and manipulation to come across as useful marketing pieces.
Though I’m reluctant to encourage the use of AI for creating videos, as it’s clearly against our business philosophy, I think there’s a niche where it makes sense. When clients come to us with a $750 budget for a video, we simply cannot accommodate their goals. Hiring professional videographers with tens of thousands of dollars in gear, having insurance for them and the company, overseeing the project creation, and every other aspect that goes into a video costs too much for us to make anything for such a low budget. In cases where there aren’t the marketing dollars to spend for a custom video, AI-generated video might be a good way to fill the lower end of the marketplace typically filled by student filmmakers and hobbyists.
I don’t see AI as a threat to our business, but rather a potential boon to smaller agencies like ours where we don’t have a large staff of editors, so maximizing the efficiency of the editing process can help us deliver high quality videos in a faster time frame than ever before. Ultimately, most of our clients want to capture interviews with important people at their companies, happy clients and customers, and footage from their businesses or of their products. Specific, custom videos require actual human beings filming other human beings and places, all under the direction of our team that arrives at the concept for each video with the marketing professionals with whom we work. Where AI starts to shine is in the post-production process.
COPYRIGHT
AI-generated videos, photos, and text cannot be copyrighted, either formally (the U.S. Copyright Office) or informally (upon creation), because copyrights are provided only to human-generated content. As a result, any video you create using AI can be taken by the competition and used for their own marketing purposes as well, without any legal recourse. For lack of a better way to phrase it, AI-generated content isn’t unique, it’s filler. For social media engagement, it could be a useful tool depending on what you’re marketing, and it is cost-effective, but what you save in cost you lose in originality and preciseness of creation. Only through custom video creation can you achieve exactly what you want, edit to your specifications, show exactly what you want to show, and have an actual marketing asset that only your company can use.
AUTO-CAPTIONING AND TRANSCRIPTION
While AI can’t create a custom video, it’s still an incredibly valuable tool for a lot of elements of video creation that have changed over the years. When we started JLB Media Productions, I personally transcribed every interview the company conducted for every video to save on costs. It took on average 2-3 hours per shoot for me to transcribe the highlights of every interview, listening to segments again and again to make sure the quotes were perfect, all so that we could use the best quotes to create an outline for each video that our editor could use to create the backbone of the narrative. Even our VP of Business Development took a crack at transcriptions for a few years, so did my girlfriend (now wife) at the time, who I hired to do the work. Everyone grew sick of it after a few years of mind-numbing transcribing. Now, we use AI to transcribe the interviews for a fraction of the cost and the same level of quality. It is, after all, not customer-facing work; it’s work done to aid the video creation process.
By the same token, AI has been instrumental in lowering the cost of transcriptions for everything from movies to marketing videos, with many services using a form of “auto-captioning” that more or less faithfully transcribes spoken language for the hearing-impaired or for phone users who watch videos but can’t hear the dialogue in busy public places. Transcription abilities have come very far since early attempts, like IBM’s ViaVoice in the 1990s, which regularly butchered the English language. I’ll never forget when my dad and I used to use ViaVoice as comedy gold as soon as he got home from work. We would start telling it a story and laugh hysterically as it got almost every word wrong. Now, AI has made transcription highly reliable and saved us from unnecessary overhead, in turn helping reduce our costs to the client.
AI TOOLS FOR VIDEO EDITING
Practically every month brings new tools for editors that use AI, but one of the best is Video AI by Topaz Labs. Their program allows for incredible upscaling of existing video footage, which a lot of independent filmmakers praise for its abilities. Footage shot in 1080p can be upscaled easily and flawlessly to 4K for instance without clear artifacts or quality degradation. Less relevant in the corporate arena than for television or film, but Topaz is capable of much more than upscaling existing footage. It can also turn lower quality, grainy footage into something usable and professional. The applications there help a company like ours with, for instance, integrating client-provided footage that may not match the professional cameras our videographers use. In a run-and-gun shooting environment, like an event for instance, where lighting may be inconsistent and thus increase the grain profile of the footage, we can use AI tools like Topaz to smooth out less-than-ideal footage (eliminating motion blur, for instance) that our videographers provide. More quality footage means more choices in the editing room and fewer risks to the client when hiring a smaller crew.
Programs like Video AI by Topaz come with a significant upfront cost and hardware demands, but can make life for an editor much easier than past decades where imperfect footage was just discarded without a second thought. Ideally, the higher quality the source footage, the better job AI can do in making it perfect, so professional quality footage shot with pro level cameras is definitely necessary for the best possible outcome. Behind the scenes, numerous programs use AI to enhance processing capabilities and tools already built into any of the major non-linear editing systems.
CONCLUSION
While the growth of AI can be both exciting and scary, depending on your viewpoint, there is more reason to be excited by the enhanced capabilities possible with AI use in video creation than there is reason for concern over it replacing jobs and destroying creativity. Ultimately, the best editors will make use of the best tools to create better, more professional videos faster than they could ten years ago. The quality of a video will still come down to the creative professionals involved from pre-production through the production crew to the final edit. While the capability to make videos entirely with AI exists, the benefits and market for those videos is primarily low-budget YouTube channels, cheap social media marketing, and niche applications that require no footage of the client’s product, services, or testimonials. For the most cost-conscious clients, such a solution might be better than no video at all, but it remains a limited use case that cannot replicate custom video production for discerning clients.