In Part 1, I went through the basic approach for iQIYI with generative AI. In this part, I’ll cover some of the use cases.
iQIYI’s Early Use Cases in GenAI
iQIYI got started by getting employees to use GenAI tools. This was a great first step and my standard recommendation for any company. Just get everyone to start using ChatGPT, Midjourney and whatever. In their daily activities. Start putting them into the workflows.
At the recent earnings report, iQIYI announced they have “integrated generative AI to advance every stage of content creation, from planning, development to promotion. AI-powered video comprehension at minute-level provides plot summaries for over 3,000 dramas.”
1. Use Cases in Marketing and Promotion
Marketing and promotion is usually the fastest adopter of Generative AI. All the big marketing agencies have started using AIGC to create ad copy, posters, slogans, graphics, etc. AIGC is just really good at creative content where there is no right or wrong (which it has trouble with).
iQIYI has said they are using AIGC in marketing for
- Improving advertising ideas and content. So, think better posters and ads. And that are done with higher productivity.
- Increasing the advertising ROI. AI is a prediction tool. Identifying which ads are going to have a greater impact is valuable.
From the CEO:
“Over the past year, we used AI for mass-producing quality marketing materials, which outperformed manual designs in click-through rates and viewership. Additionally, AI’s capability to comprehend video content at minute level enable detailed plot summaries for each playlist. Now available for over 3,000 dramas.”
2. Use Cases in Content Production
As mentioned, iQIYI already has an intelligent production system. They have mentioned they are using generative AI for:
- Extracting information from scripts.
- Scene and character breakdowns.
- Assisting in script creation and evaluation (pre-production).
- How good are the characters? Is there enough conflict in the story?
- Translation to other languages.
This sounds like text-based NLP stuff., which is the easiest place to start. The comprehension of scripts and stories can happen quickly. Then it can be used to analyze and improve scripts before production. And it can increasingly be used to write scripts. I have tried rewriting movies with ChatGPT (write a new Star Wars where Darth Vader didn’t die). They are surprisingly good.
Keep in mind, iQIYI wants to create entertainment that is high quality. And, if possible, that sets cultural trends. They want to have the kind of impact shows like Game of Thrones and House of Cards had. They want the shows that everyone is talking about. And that people buy merchandise for. That requires a lot of advanced prediction, which AI should be very good at.
But they also don’t want to be first in a genre or trend. They want to be the “only”, instead of the first.
In strategy, we would call that an adaptive strategy, where a business is good at studying and following changing customer preferences. You let other companies try lots of ideas (most of which don’t work). You see what works and then create the best products in that area. This adaptive approach is common in the fashion industry. It requires being very data driven. And the more intelligence tools you can deploy, the more adaptive you can become as a business.
Recall this matrix from BCG.
At the media event for the ‘Action Master’ season launch, Haitao Yang, Senior Vice President of iQIYI, talked about the platform’s content curation methods and consumer insights. He said success depends on a “focus on stylization by choosing the popular suspense genre, ensuring a substantial user base and high engagement.”
iQIYI appears to me to be 50% adaptive and 50% shaping strategy. But I’m not sure I’m right about that. Take that as an outsider speculating.
3. Use Cases in the User Experience
Always improving the user experience is #2 in the digital operating basics. Gathering data on customer behavior and experiences and then using that to continually improve the customer experiences. And the biggest lever to pull in this regard is personalization.
Which is exactly what iQIYI says they are focused on.
For personalization, they have mentioned using AI to:
- Improve search results.
- Improve content curation and recommendations.
- Personalizing interactive playback experiences.
Those are good early use cases. Curation and personalization are important.
But having a better understanding of what customers want in entertainment is more powerful.
And the biggest lever for a better user experience is creating better content quality.
4. Use Cases in Business Performance and Operating Efficiency
These are straight forward. And most are not specific to entertainment. AI is good at things like budget planning and resource management. So, you can keep gaining efficiencies here.
More interesting is use cases to improve monetization and other business performance metrics. Which are pretty challenging in entertainment.
***
Ok. That’s the early AI and AIGC use cases I have heard iQIYI mention. Let’s get to the bigger potential use cases. But this is mostly just me speculating.
iQIYI’s Big Potential Use Cases in GenAI
iQIYI has already been focusing on digital assets and virtual production. And this is a big idea.
Entertainment is a digital good. Like software. And you create it by filming actors and sets and such. Plus, some CGI and other digital assets. But what if create a huge library of digital assets beyond just famous characters (“we have the IP for Iron Man”). What if you turn everything that goes into creating a movie or TV show into a store digital asset:
- The key characters (which can be valuable IP).
- The secondary characters
- The extras
- The sets
- The voices
- The music
- The special effects
Here’s how they describe it on their webpage:
“To support virtual production, iQIYI has built a “Digital Asset Library” to include various assets such as 3D models, motion capture data, and panoramic videos. At the same time, iQIYI has developed its own pre-visualization software “iViz” and the “QiPai Free View System”, which includes 64 cameras and enables the fast shooting and editing of the “bullet time” effect.”
Here’s a screenshot of some of the digital assets that are available on their virtual production services page.
Then you start to use those digital assets in increasingly productions that are increasingly virtual.
Most scenes today are already part real and part virtually created. You can have a virtual production process draw assets from the asset library and combine them with the physical filming. And then some scenes can be entirely virtually created. How far are we from when entire TV shows being created virtually by drawing on an extensive library of digital assets? Animation is pretty close to this now.
From the iQIYI CEO:
“Our major original drama, The Story of Mystics, leveraged virtual production to achieve remarkable results, capturing over 3,000 minutes of footage. This pioneering technology marks a paradigm shift in filmmaking, enabling immersive, efficient, and creatively boundless storytelling for the future. Another example, for many iQIYI original productions, on-set footage could be immediately uploaded to our proprietary cloud, enabling remote production oversight and content quality checks, as well as swift real-time editing to expedite the post-production phase. This boosts our efficiency in both production and program management.”
I’m really not doing iQIYI’s virtual production plan justice. It’s really a sweeping initiative to transform the entire video production industry. Here’s how they describe it on their webpage:
“iQIYI is fully committed to the development and implementation of virtual production to promote the transformation of the industry. iQIYI Virtual Production provides comprehensive technical and production supervision services, including but not limited to: customized solutions, supplier evaluation, technical execution, process management, cost and quality control.”
That is a big idea.
- It creates a huge library of valuable IP. That can be used internally or licensed.
- It has a big impact on the cost. Digital libraries have zero marginal production costs. And they make everything reusable.
- There are big opportunities to create much better content with this process.
What happens when AIGC starts to create movies and TV shows automatically using digital assets and virtual production?
As mentioned, AI gives you better intelligence about customers and what they want. As well as what makes scripts, scenes and characters higher quality. This can increasingly be fed directly into automated virtual production.
The holy grail is the virtual production of long-form entertainment. Especially for movies and tv shows. But couldn’t this process also be used for video games? For VR? For NFTs?
And there is the other big potential idea of on-demand creation of personalized movies and tv shows. What if AIGC can create individual shows for everyone based on what that person likes the most? And where the only costs are the licenses for the digital library and the cost of compute?
That’s a lot of speculation but people are talking about this.
- Digital asset creation and management is a big use case going forward.
- Automated virtual production is another big one.
- Personalized content is pretty cool but speculative.
***
Ok. That’s it for this article. Pretty fun stuff.
Cheers, jeff
——–
Related podcasts and articles are:
From the Concept Library, concepts for this article are:
- Artificial Intelligence: Generative AI
From the Company Library, companies for this article are:
- iQIYI
———-
I write, speak and consult about how to win (and not lose) in digital strategy and transformation.
I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.
My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.
This content (articles, podcasts, website info) is not investment, legal or tax advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. This is not investment advice. Investing is risky. Do your own research.