Is AI-Generated Business Video Ready for Prime Time?
Lights! Camera! Keyboard! AI promises a new era of affordable, high-quality video productions, but is it right for your business?
Artificial intelligence is transforming video production, allowing even small businesses to create high-quality commercials, presentations, and “how-to” guides.
When professional videos can be created by simply typing a script, immediate cost savings are realized by removing lights, camera, studio, hair, makeup, and crew, says Joe Murphy, technology evangelist at AI video generator technology developer DeepBrain AI. “AI-generated videos also offer the benefit of personalizing and maintaining video content by just modifying the script and exporting the updated content,” he adds.
Traditional video editing tools are very challenging for most non-professionals, presenting steep learning curves, says Jeremy Toeman, founder and CEO of Aug X Labs, an AI-driven video technology and publishing startup. He notes that the current industry benchmark for video editing time is approximately one hour spent for one minute of output footage. “Whether the content is for entertainment, commercial, or informative purposes, it’s a slow process,” Toeman states.
While it can’t replace human creativity, generative AI accelerates video production, allowing faster and easier editing. “Generally, AI makes it easier to generate video and therefore more approachable for individuals beyond those with deep technical experience,” says Emmanuelle Rivet, a global technology leader with business advisory firm PwC.
Leading Applications
AI-generated videos can be used in the same markets as conventional videos. “In advertising, communications, entertainment, and more,” Rivet says.
There are various types of AI-generated content, observes Dor Leitman, a senior vice president with AI video technology company Connatix. “Whether it’s a video generated from information gathered automatically, human-made footage, or full video footage generated by a machine from scratch, AI-powered video can allow creators to dedicate more time to storytelling with easy-to-use creation tools,” he says.
Businesses should view AI-based video generation applications as a way “to help their workforces become more efficient while still applying human creativity and judgment to the automatic results received,” Leitman says. “AI-assisted video can enable publishers and marketers to create more diverse, engaging, creative videos at scale in a cost-effective manner,” he adds.
The Downside
AI-created video is burdened with many of the same issues that afflict other types of AI-generated content. The two most important drawbacks are related to how, and on what platform, the AI technology was trained on, says Brett Crawford, associate teaching professor at Carnegie Mellon University’s Heinz College. “Due to systemic racism, AI often generates insensitive, biased, and even racist or misogynistic content,” she warns.
Crawford also notes that organizations intrigued by AI video’s business potential should first consider the technology’s liability potential. “The AI needs to have been trained on materials not covered by copyright, so … videos generated don’t create a copyright or trademark infringement with a resulting lawsuit,” she advises.
Rivet agrees. “Under copyright and patent laws, original content needs to be used fairly,” she states. “When generating or creating a video with AI using existing materials, one needs to consider the purpose and use of the existing material in the video being created.”
Having a solid appreciation of style and aesthetics is also recommended when initiating an AI video project. Murphy warns that poorly developed AI-generated videos often look unnatural or uncanny. “The ‘uncanny valley’ creates a negative perception of the video and detracts from the message,” he says. AI-generated video should either be photorealistic or cartoonish in appearance. “The middle ground doesn’t work,” Murphy advises.
The hype currently surrounding AI-generated content in general, and AI video in particular, may fuel unrealistic expectations among potential adopters, Toeman says. “We aren’t that close to fully generative, cinematic quality video,” he explains. “We are, however, close to ‘text-to-clip’ -- more like a GIF than a video.” Toeman suggests that business leaders considering AI video projects should understand the technology’s current strengths and limitations, “which are far different from the potential future capabilities.”
The quality and relevance of AI tools and AI-produced videos will increase, Rivet predicts. “AI tools will become more abundant and powerful and costs are expected to go down,” she adds.
Future Outlook
The road ahead may not be easy, particularly as it relates to IP protection laws and their application to AI-generated video. “Regulation is still being shaped,” Rivet says. She notes that in late 2022, EU lawmakers reached an agreement on legislation to regulate AI and AI applications. Meanwhile, ChatGPT’s recent emergence has spurred EU leaders to consider tightening those rules.
For now, Rivet advises AI-generated video users to proceed with caution. “As with many emerging technologies, while it’s important to understand and experiment, businesses shouldn’t forget to apply principles of risk management, transparency, and data governance when using them,” she says.
What to Read Next:
To Meet AI/ML Needs, Leaders Need the Right Tools
Google Cloud Unveils AI Tools to Streamline Preauthorizations
About the Author
You May Also Like