Artificial Intelligence will be the “electricity” that will run the back end of the metaverse.
ChatGPT, Midjourney, Lensa AI, and deep fakes – Generative AI is transforming our content landscape. It’s prompting questions like what is art? How do AI copyrights work? Is it ethical to use generative AI for my job? Is Generative AI a fad or a useful tool for brands?
I, along with my co-founder Brandon Kaplan, Chief Innovation Officer at Journey, addressed these questions in a recent webinar we hosted about Generative AI and what it means for brands and what AI’s role is in the future of the Metavese.
Like many technologies, Generative AI is years in the making. Like how the iPhone revolutionized the consumer experience in 2006, OpenAI with it is user-friendly interfaces and APIs, will revolutionize consumer behavior in 2023 and beyond.
What is Generative AI?
Artificial intelligence is a technology that can perform tasks that traditionally only humans could perform. “Generative AI refers to artificial intelligence that can generate novel content, rather than simply analyzing or acting on existing data,” said Brandon Kaplan. If you think of artificial intelligence as a pie, generative AI is one of the slices. Robots, deep learning, machine learning, and conversational artificial intelligence are all slices of the AI pie.
The generative AI pie graph is an example of the types of AI. Some artificial intelligence may overlap with others. This is not the whole pie, especially as AI evolves and we find new ways to use it.
What is Generative AI in the context of an AI pie chart.
ChatGPT, a sub-product of OpenAI, is the most well-known AI tool. It passed 1 million users in just 5 days. “GPT3, which is the currently available version, is trained on 175 billion parameters or data points. GPT4, which is coming up this year, likely trained on over 100 trillion parameters,” said Brandon Kaplan. Generative artificial intelligence is an abundance technology. “If used ethically and in the right ways, we can use these technologies to scale what we do and create better products and experiences,” Kaplan explained. You can use it for inspiration, break writer’s block, use AI for scale. Personalize generative AI by connecting it to different data sets to produce content and reduce friction.
Types of Generative AI
What are the key areas of Generative AI that many are exploring?
Generative AI comes in all types of forms. Many generative AI uses are visual such as image, video, and avatar generation. Generative AI can also write text, summarize articles, and generate code. Audio is another area where generative AI thrives.
Text
ChatGPT is the most well-known text generative AI. It lets coders build 3x as fast. You can try it for yourself. Ask ChatGPT to generate code, review it (or ask a friend) to see how it matches up. Even though ChatGPT can generate code or text quickly, it’s important to double check it. ChatGPT is a machine learning model. While it’s trained on a large data set, there are data biases in any model.
Image
Generative image AIs can create art from scratch, modify pictures (like removing objects), or add objects to a picture. Types of image AIs include Midjourney and Lensa. Neural Radiance Fields (AI NeRF) are a new type of AI that produces 3D models from 2D pictures. NVIDIA created instant NeRFs that achieve more than 1,000x speedups in some cases. Their model requires seconds to train on a few dozen still photos (plus data on the camera angles) to render the resulting 3D scene within tens of milliseconds.
Video
Pictory and Potion video dubbing, generate videos. Generative AI can create scenes from scratch or like Runway’s Gen-1 program, transform existing videos into new ones. They call it video-to-video that can be used for storyboarding, masking, rendering and however else people unleash their creativity with generative video AI.
Audio
Generative audio AI creates media from scratch using synthetic voices. “The audio space, the ability to create audio, to convert, to edit, to dub and translate audio is going to be the underlying media element of a lot of what we all create,” said Kaplan. In the future, we might be typing into virtual keyboards, but we'll probably be doing a lot of voice.
Avatar
Avatar generation is the production of virtual beings via generative AI for fast creation and deployment of virtual and interactive characters in physical and virtual environments. Use DID or Synthesia to upload a picture of yourself or you can pick an avatar off shelf. They auto connect to ChatGPT and other tools, and you've got a fully conversation avatar potentially spun up in a matter of minutes.
The ability to have generatively designed NPCs in virtual worlds and games is earth-shattering and changes engagement in gaming platforms. But generative AI avatars is more than just looks; that's where the conversation around Cloud Edge comes into play. AWS Sim Space Weaver is one large-scale service that allows you to have almost up to a 1 million entities per instance.
Data
Synthetic data created by AI trained on rea-world data sets. “The ability to connect generative tools to some of your data and your modeling systems is going to be incredibly exciting,” said Kaplan. Use generative AI to make variable data sets based on a specific volume of data you already have on your consumers or audience. You can also use generative AI to clean, filter, identify, and fill in data and even explore new insights. Use generative AI to model data like audience behavior, product design, and physical retail environments.
What Generative AI Means For Brands
Hyper personalization
These tools will allow us to personalize a lot of how we receive content throughout the world and in our consumer experiences. “You can generate, predictive content on live data that's going on in the world, and there there's a ton of different ways that you can personalize content out into the world,” said Kaplan.
One application is retail. Kaplan explained, “people all looking at the same main image, the same backgrounds, the same 5 or 6 images. They're seeing the same captions, the same pricing. They've seen the same descriptions.” Generative AI offers the ability to customize websites, products and descriptions for people looking at the page. Instead of having to guess at a size chart for clothing, the AI will automatically display the page based on what works for them. This will lead to more conversions.
Consumer Choice in Customization
Next generation advertisements are another area where generative AI can help. “People are all looking at the same main image, the same backgrounds, the same 5 or 6 images. They're seeing the same captions, the same pricing. They've seen the same descriptions,” said Kaplan. He recommends you feed your data through a generative AI to subtly deliver different versions of an ad that speaks to people where they are. While there is a hype cycle that generative AI will go through, these are tools that companies and brands can start using today to figure out revenue streams and find audiences in ways that work for them.
Don’t Trust Everything You See or Read
ChaptGPT is only a few months old. Its data set only goes up to 2021. Kaplan warned, “There's a lot of hype. There are a lot of experts. So, tread lightly. Educate yourself, but don't trust everything you see.”
Generative AI in 2023 and Beyond
We are just beginning to see how generative AI will be used at work, in play, and how it will impact our jobs. Generative AI will have a large impact on the creator economy and how brands interact with their communities. Generative AI is one more tool that brands should add to their stack. As long as brands do their own research, have an open mind, and are not afraid to experiment we believe generative AI will be a success.
Are you ready to use generative AI for your brand? Do you still have a lot of unanswered questions? So many new questions will arise as the use of AI continues to progress and its adoption continues on the fast track.
As I wrote here in Forbes in an article back in 2020, when we turn on the lights or open the refrigerator, we know there’s electricity powering them but we don’t think about how it works. In the same way, artificial intelligence, sensors, and robot-to-device communication will be the “electricity” that will run the back end of the metaverse. The physical world we live in becomes machine-readable, clickable, and searchable. There will be new interfaces and new ways to navigate and create content. We will create new lexicons and even new architectures for our digital selves and the digital world around us.
Written in collaboration with Lily Snyder.
To stay on top of the latest on the Emerging Tech, Metaverse, Web3, connected luxury fashion and other new and emerging business and tech trends, make sure to follow me on LinkedIn, Instagram, and YouTube, and check out my new book Into the Metaverse: The Essential Guide to the Business Opportunities of the Web3 Era.
