In the dynamic world of games, artificial intelligence (AI) is transforming the landscape, offering exciting possibilities for enhanced gameplay experiences. Communicating AI agents have emerged as a focal point, raising questions about their value and potential benefits.
Let’s delve into the realm of AI in gaming, including such diverse aspects as the added value of communicating AI agents, the delicate balance of emotional recognition, the characteristics of large language models (LLMs), the role of generative AI in orchestrating narratives and the future of scriptwriters in this evolving landscape.
Unveiling the added value of communicating AI agents
To truly understand the value of communicating AI agents, it’s essential to distinguish what we mean by ‘AI agents’ and ‘virtual assistants’. From the perspective of an AI developer, an AI agent refers to a system that autonomously finds solutions to its tasks and goals, using various methods without being under direct human control.
As an example, an AI agent’s goal could be to find solution strategies for a difficult situation in a game. In doing so, an AI agent could search the web for walkthroughs, analyse videos of human players using computer vision, try to understand the user interface and consolidate different sources to derive an optimal solution approach.
Initially, this has nothing to do with a virtual 3D character or assistant as we often imagine. However, to make an AI agent’s capabilities more accessible, they could be driven through a natural language interface and be represented by a virtual character.
Serving the players’ needs
By harnessing generative AI, these agents could create hyper-personalised in-game items and offer guidance tailored to individual players. Imagine an NPC (non-player character) shop owner crafting unique items based on your desires. Or an Xbox AI assistant providing expert advice for games like Minecraft and Fortnite, eliminating the need for extensive YouTube searches that lead to the player leaving the game. Moreover, AI has the potential to fill the void left by frustrated players who prematurely exit multiplayer game sessions, paving the way for a seamless and engaging experience.
It will be crucial to identify real added value for users and not just enrich old game design approaches with a little AI. It’s important to understand that AI, in this sense, can offer much more than just an NPC as a more or less natural conversational partner in a role-playing game. NPCs as endlessly talkative quest givers are a dead end.
LLMs: enriching gameplay through conversation
So, what does generate added value? When a game adapts better to the player’s needs or even creates entirely new possibilities. There can be very interesting game concepts where the focus is on conversing with AI, breaking the fourth wall (directly addressing the user rather than their in-game character as part of the storytelling). Just think of The Stanley Parable, a game that seems to understand the player’s actions and reacts accordingly.
It’s also essential to provide the AI with a degree of priming or a specific task to ensure meaningful interactions and avoid arbitrariness. By striking the right balance, large language models can truly enhance the gameplay experience and unlock uncharted gaming territories. This has huge potential for exploring rich worlds or extensive topics – like Niantic created with their Black Forest educational experience Wol.
The intricacies of emotional recognition and players’ trust
As AI advances, its ability to analyse players’ emotions becomes increasingly feasible, whether through gameplay or interactions with non-player characters. This opens the door to AI agents offering in-game items as a solution to frustration. However, the implications of emotional recognition raise concerns about user manipulation and trust.
We must be aware that emotional manipulation has always happened in entertainment. When we cry during a movie or shout at the TV after losing a boss fight, these are intentionally triggered emotional states. While emotional manipulation has long been a part of entertainment, developers must tread carefully. Striking a balance between leveraging AI’s capabilities for the benefit of players and considering ethical implications is crucial.
Gaining and maintaining players’ trust in an era of AI-driven emotional recognition is vital. Above all, developers and designers must be aware of this responsibility and consider ethical perspectives, not just commercial ones.
With the almost magical possibilities of AI, and increasingly XR (extended reality) as a medium, gaining users’ trust and responsibly handling these fantastic opportunities becomes crucial. Industry leaders recognise the significance of responsible implementation and ethical considerations, which has prompted a shift in their strategies and positioning in recent years.
Generative AI enhances game production
As AI technology evolves, the question arises: does the true strength of generative AI lie in orchestrating the overall game narrative rather than in individual dialogues? While the vision of an AI game master or designer who autonomously orchestrates an entire game is yet to be realised, generative AI offers different opportunities. Maybe conversational AI agents could serve game developers even better than they do gamers.
This way, LLMs can serve as an orchestration layer for the AI agents that autonomously find solutions for specific tasks. It will likely be possible to create simple games almost entirely with the right AI prompts. On the one hand, with a good generative AI production pipeline, it would surely be possible to have an autonomous game production factory to churn out hidden object games. On the other hand, generative AI is still finding its place in more complex game productions.
Legal implications are still unresolved
More and more elements of the traditional 3D pipeline can be implemented more efficiently with the help of AI. Currently, this is already extensively used in the conceptual phase. However, uncertainties, such as copyright concerns, still surround the utilisation of AI-generated assets. The ongoing debate on copyright protection for AI-generated materials poses challenges for publishers and developers.
Pragmatic solutions, like studios training their own base models, will likely emerge to ensure the protection of intellectual property. At the same time, concerned employees might not be willing to sign off on this new paradigm. These new methods will empower smaller studios, who are less risk-averse, to shake up the market quite a bit, as they will have more possibilities with smaller budgets.
Artists and scriptwriters in the age of AI: adapting to new realities
The evolution of AI raises concerns among scriptwriters and graphic artists, reminiscent of the shifts experienced in stock and virtual photography. While AI-generated content poses a potential challenge, it’s unlikely to render scriptwriters obsolete in the near future. However, they must differentiate their creative work from what AI can effortlessly produce. Instead of fearing obsolescence, exploring new directions and embracing AI as a creative tool can lead to innovative outcomes for artists as well.
Furthermore, demands for the mandatory attribution of AI-generated texts face significant challenges due to the abstract nature of texts and the difficulty of distinguishing AI-generated content. Although the situation may differ slightly for images and 3D content, similar hurdles exist. Incorporating transparency and traceability into AI tools, revealing the influences and origins of generated works, poses both technical and commercial challenges that require careful consideration.
Embracing AI for a new era of gaming
As AI becomes an increasingly integral part of the gaming landscape, its potential for transforming gameplay, narratives and the creative process is evident. While challenges and ethical considerations persist, developers, designers and players alike must navigate this evolving terrain and embrace responsible AI implementation to unlock the full potential of gaming experiences.
By harnessing the power of communicating AI agents, large language models and generative AI, the future of gaming holds exciting possibilities to push boundaries and reshape the industry for years to come.