The Future of Business Video: Better Communication with GenAI

Before I delve into Generative AI and video, let’s look at the love-hate relationship we as a species love watching others but rarely ourselves in video. This is probably one of the biggest inhibitors to people embracing content creation.

I often say,” I have a face for radio and a voice for print.” But that’s not just me; that’s part of how many of us feel when we watch or listen to ourselves in video. This is caused by an interplay between what we believe and perceive, which lies in the science of our discomfort with our own recorded image and voice.

Our bodies and brains conspire uniquely: while we speak, our voices resonate through our skulls, creating a deep and familiar sound through bone conduction. But, when we play back a recording, we hear a stranger version that's traveled only through the air. Our brain, which relishes the comfort of the known, is taken aback by this unfamiliar sound, sparking a cascade of emotional and cognitive reactions. It's like expecting to sip a hot cup of coffee only to find it's gone cold. In video, we're used to seeing ourselves in the mirror, a reversed image. Seeing the non-reversed version of ourselves in videos can feel unfamiliar and off-putting.

These are among the top reasons why we tend not to like videos of ourselves, but despite that, there are many other ways to use Generative AI to create business videos that don’t require us to sit in front of our camera as a talking head.

Video for Business

Video is becoming increasingly crucial for businesses to convey their product value—the total time spent watching videos increased by around 249% in 5 years from 2015 to 2020. In addition, short-form video is the top leveraged media format in marketers’ content strategies. (HubSpot Marketing Trends Report, 2024). Fifty-one percent of marketing professionals globally also consider video as the content that delivers the best ROI

Also, video isn’t just about marketing. It’s about streamlining communications and reducing operational costs like support calls. It can be a much more effective way to convey how to use or evaluate products than written content.

In their 2023 Digital Media Trends: Immersed and Connected report, Deloitte notes virtually every age segment shows a greater affinity for user-generated content (UGC). 

Trends in Video for Business

YouTube has over 38 million active channels, and 15 million channel owners are content creators. Over 22,000 channels have 1 million subscribers. Wyzowl, an animated video production studio, does an annual survey on video marketing trends. Ninety-one percent of businesses used video as a marketing tool in 2023. However, 30% of people lack the time to create video content, and 18% don’t know where to start. Also, making videos has not been cost-effective for many; over 33% of videos cost over $1000.

GenAI for Video Creation: Why It's the Move Productive Choice

Generative AI (GenAI) for video production has presented a transformative alternative to traditional recording methods. GenAI is a revolutionary approach that empowers creators with efficiency, creativity, and many other benefits.

• Cost-Effective & Time-Saving: Traditional video production can be a costly and time-consuming endeavor involving equipment, actors, and various logistics. GenAI dramatically reduces these expenses and processes, allowing for rapid creation and iteration.

• Unlimited Creative Potential: GenAI opens up a world of endless possibilities where the only limit is your imagination. It enables the creation of scenarios and visuals that might be impractical or impossible to film traditionally.

• Accessibility & Inclusivity: GenAI democratizes video production, making it accessible to individuals and businesses without the resources for high-end video shoots. It ensures that quality content isn't the exclusive domain of those with deep pockets.

• Consistency & Customization: Maintaining consistent quality and style across videos is effortless with GenAI. It also allows easy customization to target different segments or adapt to various platforms.

• Safety & Feasibility: Some scenes are risky, sensitive, or unfeasible to shoot in real life. GenAI eliminates these concerns, offering a safe environment to bring any vision to life.

Six GenAI Tools for Creating Videos

Several high-quality tools can make videos faster and more efficiently than conventional methods.

1. RunwayML: A user-friendly platform that offers powerful video editing and generation features. Its intuitive interface makes it a favorite among beginners and professionals alike. You may recall the rock world if you saw the Academy Award-winning movie Everything Everywhere All At Once. This animation wasn’t filmed using stop footage. It was 100% generated by RunwayML.

2. Synthesia: Renowned for its ability to create realistic video avatars and deep fakes, Synthesia is perfect for creating engaging talking-head videos and presentations without needing a real actor.

3. DeepBrain AI: Specializes in creating AI human models that can be used for various video applications, offering lifelike interactions and presentations.

4. Descript: An all-in-one video editing tool that leverages AI for tasks like auto-transcription, overdubbing, and editing, making it a powerful tool for creators of all skill levels.

5. Pictory: Ideal for turning long-form content into engaging video summaries. It uses AI to identify key moments and create concise, visually appealing videos.

6. Lumen5: Leverages AI to help users transform text content into video. It benefits marketers and content creators looking to repurpose blog posts or articles into dynamic videos.

Bonus Application:

Leonardo: While Leonardo is more of an image tool, I like its Image2Motion feature to take images and give them simple movement.

Google Research VideoPoet

Just before Christmas, Google Research recently announced an impressive modeling approach called VideoPoet. This approach can easily convert autoregressive language models into high-quality video generators.

VideoPoet uses a unified vocabulary to blend various modalities, such as images, video, and audio, seamlessly. It can handle tasks such as text-to-video, image-to-video, and more by incorporating multimodal generative learning objectives. The model's zero-shot ability, like text-to-audio, adds an exciting dimension to its capabilities. VideoPoet breaks traditional boundaries and offers a powerful tool for innovative and engaging multimodal content generation.

Although the company has not disclosed any plans to make the model available, it may be integrated into a planned Bard Advanced at some point. You can find more examples in full resolution on the VideoPoet project page.

Use Cases for GenAI-Created Videos

Video content has become a driving force across nearly every industry, allowing brands to connect with consumers in engaging and memorable ways. With advances in generative AI, creating customized video content is now quicker and more accessible than ever before.

GenAI empowers any organization to produce high-quality, personalized videos at scale for a fraction of traditional production costs. From advertising campaigns to corporate communications, GenAI unlocks a breadth of video use cases with the ability to generate footage tailored to any audience or purpose instantly. The potential applications of this transformative technology include:

• Marketing & Advertising: Create personalized and targeted ad campaigns or product demonstrations without extensive shoots.

• Education & Training: Develop educational content, tutorials, and training videos that are both engaging and informative.

• Entertainment & Storytelling: Produce films, music videos, and other creative projects with unprecedented visual effects and narratives.

• News & Reporting: Generate news segments with virtual anchors, allowing for timely and cost-effective news delivery.

• Corporate Communication: Enhance internal communication with personalized messages and presentations from virtual leaders.

• Social Media Content: Quickly generate trendy, platform-specific content to engage with audiences across various social media channels.

Unlocking Creative Potential with Generative AI in Video

The emergence of new AI tools that generate high-quality synthetic video has exciting possibilities but raises ethical questions. On the positive side, these generative video AIs may allow more people to craft visual narratives and share diverse perspectives through an engaging medium by automating technical aspects of production.

However, as realism improves, the ability to detect fake AI-generated video could become problematic. This risks enabling the propagation of misinformation or media manipulation. There are also potential economic impacts as traditional video production roles are disrupted.

Maintaining human oversight and transparency around AI use will be crucial as we progress. We need continued discussion to establish ethical guardrails that prevent harm. If carefully navigated, democratizing creative expression through such technologies could open new horizons.

But we must weigh both the benefits and risks as capabilities advance. Striking the right balance presents a crucial societal challenge regarding the future of generative video AI.

Prompt of the Week: Create a Video Script using ChatGPT

When I use the tools I mentioned above, I typically use ChatGPT to create a script and then use that script with Synthesia to generate videos. Here’s a simple example that you can adjust for your needs.

Here are some guidelines to help. Typical conversations move along at about 120 to 150 words per minute. This is a good baseline for how many words to write into each segment of your video script because you want your speech to be conversational.

Replace the [TOPIC] with the topic of your choice.

You act as a very proficient SEO and high-end copy writer that speaks and writes fluently in English.  Provide the content for a video on [TOPIC].   

There must be an intro, a body, and a conclusion.  

Body must contain 5 segments, having each segment must be of 200 words length.   

Script length must be a minimum of 840 words.  

Of course, you could also use this to create TikTok or YouTube scripts.