2024 marks a year of explosive growth for global AI applications. The latest data from Sensor Tower indicates that the total revenue generated by AI apps globally has surged by an impressive 51% in the first eight months, compared with the corresponding last year. This trend forecasts a full-year revenue of approximately USD 3.3 billion for 2024, underscoring the undeniable value and potential of the AI market for developers across the globe.
At the 2024 Gamescom Congress, Hilary Mason, CEO of Hidden Door, delivered a keynote address. The event drew top experts from gaming, culture, and education worldwide to explore upcoming trends and the trajectory of digital entertainment.
Hilary Mason
Co-Founder and CEO of Hidden Door from the USA
Topic of the speech: The Joys of AI
Hilary Mason, co-founder of AIGC gaming startup Hidden Door, harnesses creative works like renowned books and film IPs to train machine learning models, offering players unique multi-IP role-playing game experiences. In the presentation, Mason discussed the opportunities and risks of machine learning systems and showcased pragmatic approaches to AIGC in technology, product, and business applications.
01/ AI's Ascent: Generating, Predicting, and Emulating
With 2024 on the horizon, AI has weathered multiple cycles of hype and skepticism. We now confront the critical choice of employing this technology responsibly. It promises unprecedented personalization, enabling innovative interactions with the content we adore. However, without prudent use, there's the risk of being reduced to a cost-cutting tool or worse, a violator of intellectual property.
Mason noted that there are still many misconceptions about AI in today's society. Although many in the audience have used AI to some extent, the definition and expectations of the technology remain vague. She emphasized that AI is not just data science or big data, nor is it solely machine learning; it's a technology capable of generating content, predicting outcomes, and emulating human behavior.
02/ AI product development experience in the early days
Long before the current "AI boom," Mason was deeply involved in developing artificial intelligence projects. In 2009, while teaching machine learning, she was frequently asked the same questions, prompting her to write a script for automated responses. Although the scripts performed well in most cases, unexpected issues arose when replying to follow-up emails. This episode underscored for Mason the ongoing need for human oversight to address AI's unforeseen shortcomings, highlighting the limitations despite its advancing power.
In 2014, AI pioneer Mason launched Fast Forward Labs, pioneering an NLG system designed for real estate advertising. The inspiration behind this initiative was a virtual boyfriend app that faced removal over issues of misconduct with young users. Driven by curiosity about the models' language-generating potential, Mason is now dedicated to exploring how to harness these technologies responsibly, without perpetuating harm.
In 2015, Mason and her team delved into the realm of image analysis, crafting a deep learning model to identify image contents. Their model showed promise in recognizing “hamburgers”, but faltered with “french fries” and “water”. This challenge prompted Mason to conduct a thorough critical assessment of AI, spurring her to consistently pursue enhancements and optimizations in the technology.
03/ The Essence and Limitation of Generative AI
Mason stresses the vital importance of truly understanding AI to harness it effectively. A key feature of today's AI is its ability to generate content, but these systems merely mimic the patterns in their training data, often leading to 'generic' outputs lacking in originality. Case in point: when MidJourney and ChatGPT were tasked with creating a 'live Gamescom audience photo,' the images failed to reflect the diversity in age, gender, and race. Mason warns that such models can perpetuate and even exacerbate the biases inherent in their training data, generating content that may be both incomplete and detrimental.
Additionally, Mason also noted that generative AI doesn't grasp the truth; it just generates plausible language or images from data patterns, without verifying factual accuracy. This is beneficial for entertainment and gaming due to the diversity it allows, but it can be problematic in fields that require precision and authenticity.
04/ The application of AI in the gaming field
Mason considered that AI cannot simply reduce costs, but change the work procedure, bringing new creativity and possibility in gaming development. AI helps game developers build the fragments and details of stories, making them more like directors, not writers while players can start their game experience created by themselves in these fragments. The revolution will make the game world more interactive and changeable, which players can produce a bigger influence to their game experience.
.At “Hidden Door”, AI offers diverse immersive role-playing, letting users craft characters within the realms of their favorite fiction, film, or TV series. Mason highlighted that their model can process and translate each player's input into coherent English, broadening the game's reach. An example is the system's ability to automatically render Russian inputs into English, ensuring inclusivity for a wider player base.
However, Mason emphasizes that AI, while it plays a significant role in content creation and improved gaming experiences, still cannot replicate human creativity. She asserts that genuinely innovative concepts stem from human minds, not machines. Hence, the key challenge in AI's progression is how to use it to amplify human creativity, not to replace it.
05/ The Ethics of AI and Corporate Responsibility
AI's application poses a host of ethical and legal quandaries, notably around data usage and intellectual property. Mason observes that AI models often rely on large datasets without explicit permissions, leading to legal challenges. She insists that corporations should integrate ethics into their AI Product Promotion strategies, carefully designing products and business models that reflect social responsibilities.
Beyond efficiency gains, AI's deployment should aim to unlock new potential. If a technology can achieve what was once impossible at a lower cost, the focus should be on leveraging these tools for innovation, beyond merely cutting costs.
NetMarvel is a global one-stop performance marketing platform, dedicated to providing mobile game and app advertisers with solutions for user growth, advertising marketing, and promotional campaigns, helping to boost business growth.