AI-generated and AI-driven characters are still in their early stages of development, with significant progress expected before they achieve full maturity. Nevertheless, they are already exerting a noticeable influence on our daily lives, and their utilization is steadily increasing.
These characters assist us in various tasks such as navigating cities, managing our smart homes, organizing schedules, engaging in conversations on website support platforms and social media platforms, among other functions. Moreover, developers within the field are enthusiastic about the prospects of these characters being integrated into augmented, mixed, and virtual reality environments, heralding a new era of human-computer interaction.
However, the current status of AI-generated characters raises questions: What functionalities can they presently perform, and what limitations do they still possess? Furthermore, what ethical and practical concerns should we consider regarding their deployment? Lastly, what advancements can we anticipate in the near future? To delve into these inquiries, we have enlisted the insights of three industry experts to shed light on the evolving landscape of this burgeoning industry.
A quest for sentience
Jacob Loewenstein, head of business at AR startup Spatial, describes the endeavor of creating AI characters as an aspiration to imbue them with responses akin to sentient beings, resembling living organisms or even humans.
The prospect of achieving such apparent sentience seems tantalizingly close to fruition. Presently, some AI characters possess capabilities like recognizing facial expressions and various objects, engaging in spoken dialogue, answering inquiries, and generating lifelike outputs by processing diverse inputs.
When provided with a digital embodiment, these characters can navigate virtual environments through headsets or screens. However, they currently struggle to comprehend human intent and analyze sentiments beyond surface-level understanding.
"We're still lacking in that aspect," acknowledges Loewenstein. While these characters can identify shapes and outlines, grasping intent and sentiment proves considerably challenging.
Nonetheless, Loewenstein expresses optimism that with sufficient time and data, advancements will occur. Certain companies have already made strides in addressing aspects of this challenge. For instance, Affectiva employs facial cues and voice analysis to enhance driver safety by detecting emotional and cognitive states. Similarly, Spirit AI utilizes its character engine authoring tool and SDK to create an AI capable of monitoring abuse, identifying harassment, and fostering a safe online community environment.
Greg Cross, co-founder of New Zealand startup Soul Machines, highlights their autonomous "digital humans," equipped with a cognition engine that mirrors human nervous system responses to interpersonal interactions with minimal latency. These AIs react to gestures such as smiles and yawns, striving to discern contextual cues and adjust their behavior accordingly.
Soul Machines leverages these digital humans to revolutionize customer experiences for clients like Mercedes Benz, Autodesk, and IBM Watson. Initial results are promising, with digital humans emerging as the preferred interface for customer service engagements. This trend aligns with a recent Salesforce report indicating that 69 percent of people prefer chatbots for swift brand communication.
Despite progress, considerable challenges persist. AI agents often struggle with the nuances of facial expressions, speech patterns, and body language. Even with some degree of face, vision, and emotion recognition, conversations with chatbots and AI assistants typically rely on human-scripted dialogue.
While demonstrations like Google Duplex showcase impressive capabilities, AI agents still grapple with generating dynamic conversational content. Interactions with AI assistants such as Cortana or Alexa often reveal limitations beyond basic tasks like information retrieval and appointment scheduling.
In the realm of AI-generated characters, the focus has shifted from visual fidelity to behavioral realism and responsiveness, especially in VR and AR environments. Many AI characters exhibit minimal sentience, operating within predefined routines and basic recognition systems. However, the evolving landscape suggests a departure from this status quo in the near future.
Reshaping interactive entertainment, slowly
According to Spirit AI's character AI specialist Aaron Reed, the domain of video games, where advancements in AI characters have often lagged behind graphics and animation processing, is witnessing a gradual influx of innovation from various fronts.
In titles like Shadow of Mordor, for example, procedurally generated orcs evolve based on player interactions, remembering previous encounters and adapting behaviors accordingly. Similarly, the Forza series utilizes learning neural networks to simulate AI drivers reflecting players' driving styles, enhancing the illusion of competing against human opponents globally.
Indie games such as Façade and Prom Week delve into sophisticated AI characters that dynamically respond to player input, provided players engage with the game's narrative dynamics. Meanwhile, Event[0] challenges players to interact with a potentially self-aware computer intelligence through text input.
Games like Crusader Kings and Dwarf Fortress employ algorithmic systems and procedurally generated family trees to generate AI characters with emergent behaviors based on value decisions.
However, despite these advancements, Reed notes the absence of a breakout game centered around an AI-driven character in a narrative context, echoing a broader trend in entertainment and media. While digital characters like Lil Miquela and Hatsune Miku enjoy substantial followings, they lack true AI agency, being controlled by marketers rather than algorithms.
Nonetheless, Reed anticipates an indie game featuring an AI-generated character to make a significant impact in the near future. He suggests that triple-A studios, while traditionally cautious about embracing complex procedural systems and AI-driven narratives, are beginning to explore such possibilities, particularly as text-to-speech technology improves and natural language processing evolves.
This potential paradigm shift could redefine gaming experiences, emphasizing character interactions and empathy over traditional gameplay mechanics. Reed highlights a Spirit AI VR demo where AI characters respond to players' body language, adapting their behavior accordingly, hinting at a future of immersive and performative storytelling.
As authoring tools become more accessible and developers refine the craft of AI-driven storytelling, Reed anticipates exciting developments on the horizon, inviting a new era of player engagement and narrative exploration in gaming.
The AI field has a diversity problem
Several issues need addressing as AI progresses. One significant challenge is the lack of diversity within the field, a problem that Loewenstein highlights.
Despite a growing acknowledgment of the importance of inclusivity, there remains a disparity between developers' current practices and their desired goals. This gap is evident in various aspects of AI development.
For instance, animation libraries for AI characters heavily rely on motion capture data predominantly sourced from white male developers and a select group of professional actors. Consequently, the range of movements and gestures exhibited by AI characters in virtual, augmented, and mixed reality environments remains limited.
Moreover, biases persist in other aspects such as speech patterns and decision-making processes, often reflecting subtle prejudices toward white men. Even efforts to train AI models on diverse datasets may not fully eradicate underlying biases embedded in the code.
While the democratization of AI character creation allows participation from a broader audience, the lack of diversity among developers restricts the diversity of experiences achievable with AI characters.
Recent instances, such as Amazon's biased AI hiring tool and Microsoft's controversial Twitter bot Tay turning racist, underscore the ramifications of unchecked biases in AI systems. Additionally, emerging technologies like deep fakes and voice mimicry pose new challenges, enabling the creation of deceptive content using public figures' likeness or voice.
Although companies like Truepic are developing methods to detect such manipulations, this could trigger an arms race in deception and countermeasures. Loewenstein anticipates that the advancement of this technology will necessitate proactive measures to address its ethical implications long before fully autonomous sentient characters become a reality.
Ethical dilemmas
The ethical considerations surrounding the development and deployment of AI characters are gaining prominence with each passing day. Many experts within the industry emphasize the need for a discourse on the rights and wrongs of AI characters, particularly as the technology approaches potential breakthroughs.
"We're reaching a stage where interactions with entities that appear human but are not truly human are increasingly common," notes Reed. Legislators are taking notice as well; a new California bill mandates that commercial and political bots disclose their non-human identity.
The proliferation of AI characters also raises significant privacy apprehensions. Placing AI characters in surveillance roles can obscure the notion of being observed and having data collected about oneself. Interacting with these characters often entails divulging contextual, emotional, and personality-related information.
The ethical implications of data collection by AI creators are paramount. Some, like Soul Machines, advocate for end-user ownership and control of collected data to ensure accountability and transparency.
Recognizing the need for proactive measures, individuals in the AI community are initiating projects such as The Ethics and Governance of Artificial Intelligence Initiative. These initiatives aim to research, discuss, and establish agreements on ethical dilemmas surrounding AI. However, the landscape remains nascent, lacking universally agreed-upon manifestos or self-regulation systems.
Looking ahead
Loewenstein observes a growing acceptance of integrating AI characters into daily life, particularly evident in children's interactions but also increasingly embraced by adults using AI assistants on their devices like Amazon Alexa and Google Home. "If it enhances utility and simplifies life, people will adapt comfortably," he remarks.
Regarding the potential of AI-generated characters to supplant humans in customer service, Reed suggests that writers won't be replaced anytime soon. Instead, AI technology augments writers' capabilities, enabling them to create content more efficiently and effectively.
This enhancement process, however, will require time to mature. Reed draws a parallel with the adoption of CD-ROM technology, highlighting the interval it took for innovative applications like Myst to fully leverage its capabilities.
AI-generated characters represent a novel approach to storytelling, offering fresh perspectives on the human experience. Reed anticipates a prolonged period of experimentation across various software domains, including games, education, and entertainment. He eagerly awaits the insights that future AI characters, akin to "the Nabokov's of AI characters," will offer on humanity.
In the interim, the prevalence of AI-generated characters is expected to rise, with significant advancements in dynamic conversational capabilities anticipated within the next couple of years. With our world increasingly digitized and interconnected, Cross envisions a surge in demand for AI characters that entertain, assist, and reassure us in various settings, including self-driving cars, where trust is paramount. As AI characters become more human-like in appearance, behavior, and speech, trust in them is likely to deepen.
However, the proliferation of AI characters won't be limited to a few instances; it's projected to expand exponentially, with hundreds, if not thousands, serving different brands and purposes, potentially outnumbering humans in the future.
Conclusion
the emergence of AI-generated characters heralds a transformative era in technology and storytelling. As these characters become increasingly integrated into our daily lives, they offer a glimpse into the future of human-computer interaction. From enhancing customer service experiences to redefining entertainment and education, AI-generated characters hold immense potential to shape various aspects of society. However, alongside their promise come significant ethical considerations, including privacy concerns and the need for diversity and inclusivity in their development. As researchers, developers, and policymakers navigate these complexities, the evolution of AI characters will continue to unfold, paving the way for new narratives and experiences. With ongoing advancements in artificial intelligence and natural language processing, the horizon for AI-generated characters is boundless. As we journey into this digital frontier, the possibilities for innovation and exploration are limited only by our imagination and our commitment to ethical stewardship in the development and deployment of these remarkable creations.
Other related Articles
Video games are in for quite a trip’: How generative AI could radically reshape gaming
CNN — When you read “artificial intelligence,” you might think of new innovations such as Chat GPT, but AI has been used in video games since the 1950s. Read full article here.
Disclosure
Disclaimer: gameProXL is an Amazon Affiliate. Pages on this website may contain Affiliate links that will redirect you to other websites that allow gameProXL to earn a small commission if you make a purchase (at no additional cost to you). See full website Disclaimer.