I began exploring the differences between generative AI and traditional AI. The key distinction lies in their capabilities and applications.
Traditional AI primarily focuses on analyzing data and making predictions. In contrast, generative AI creates new data that resembles the data it was trained on. To put it simply, traditional AI excels at recognizing patterns, while generative AI is skilled at producing new patterns. While traditional AI can analyze data and explain its findings, generative AI can use the same data to generate something entirely new (Marr, 2023).
AI in video games helped me understand where it is useful. Classic AI helps non-player characters (NPCs) move around the game world. Combat AI allows enemies to make basic choices, like attacking, fleeing, or defending, based on what players do. This can be seen in games like Halo and F.E.A.R.
Procedural content generation (PCG) creates different environments automatically, which makes games more replayable, as shown in Minecraft and No Man’s Sky.
Cheating AI gives NPCs some advantages to balance the game. For example, it may speed up opponents in racing games (e.g., Mario Kart) or reveal players’ locations to enemies, keeping players engaged (Bardon, 2024).
Generative AI helps speed up game development with tools like Artbreeder and GANverse3D, which efficiently create visuals and 3D models. AI also improves voice acting by using text-to-speech systems that can mimic various accents and emotions. Dynamic content generation creates responsive terrain, dialogue, music, and quests, enhancing the gameplay experience. Moreover, automation tools like GameDriver and Modl.ai improve quality assurance by simulating playthroughs. This helps quickly find bugs and balance issues.
This made me think about how Generative AI can create highly personalized content that keeps players engaged for longer. This raises concerns about the risk of addiction or exploitation.
One of the most striking real-world examples of generative AI’s immersive power is the Mantella mod for The Elder Scrolls V: Skyrim. This mod transforms over 3,000 NPCs into dynamic, responsive characters capable of holding real-time, natural conversations using speech-to-text, large language models, and text-to-speech synthesis.
These AI-powered characters remember your past interactions, perceive in-game events and objects, and can even make independent decisions, like choosing to fight you or follow you. With Mantella, you can flirt with Serana, berate Maven Black-Briar, or get into philosophical debates with Brynjolf. It’s no longer a scripted experience. When AI characters can see, think, remember, and respond like people, we’re no longer just playing a game.
This reminded me of Nozick’s experience machine thought experiment (EMTE)…
In The Examined Life (1989), Nozick presented the EMTE as follows:
Imagine a machine that can give you any experience you want. When you connect to this machine, you can feel what it’s like to write a great poem, create world peace, or love someone and be loved back. You can enjoy these feelings completely. You can choose your experiences for tomorrow, this week, this year, or even for the rest of your life. If you have trouble thinking of experiences, you can use ideas from biographies or stories written by novelists and psychologists. You can live out your biggest dreams in your mind.
Would you want to do this for the rest of your life? If not, why not? Other people can use these machines too, and they’re provided by friendly beings from another galaxy, so you don’t have to refuse to help others. The question is not whether to try the machine for a short time, but whether to stay connected for life. Once you enter, you won’t remember that you chose this, so your pleasures will feel real and won’t be affected by knowing they are from a machine. You can also choose to include uncertainty by using the machine’s option for randomness, which lets you pick from different choices.
I had to discuss this in my philosophy class from my first year of college, and pretty much everyone said no, they wouldn’t do it. But just because one class says “no,” doesn’t mean the rest of the world’s population agrees.
In today’s world, there are individuals known as “shifters” who believe they can transfer their consciousness to an alternate reality. Shifting is a phenomenon, especially popularized online, where individuals claim to move their consciousness to a desired alternate reality, often fictional ones like Harry Potter or anime worlds, through mental scripting and meditation. The mechanics of this process are complex, but my main point is that some people are so dissatisfied with their current lives—or simply yearn for more—that they attempt to leave their present existence behind. They seek to escape into a life that aligns perfectly with their desires and aspirations.
What once seemed metaphysical now has a technological counterpart. Now, consider giving aspiring shifters an experience machine. It requires none of the effort to “shift.” It sounds surreal and futuristic; how could a machine know precisely how you feel and what experience should be generated for you?
The answer, however, lies with Elon Musk and Neuralink. Right now, the Neuralink website states:
“We are currently focused on giving people with quadriplegia the ability to control their computers and mobile devices with their thoughts,” and, “In the future, we hope to restore capabilities such as vision, motor function, and speech, and eventually expand how we experience the world.”
I’m not going to lie and say that this is inherently bad; it sounds incredibly beneficial for those with disabilities! However, consider Ozempic and Wegovy: originally intended for individuals with diabetes, they are now increasingly used for weight loss, particularly among those with financial means. How long will it take before Neuralink is used for shits and giggles by those more wealthy than people that would actually benefit from it?
Consider this: by combining Neuralink with generative AI, we could create an experience machine. It’s a possibility, not a certainty.
Movies like Don’t Worry Darling, The Matrix, WALL-E, and Ready Player One offer cultural previews of this very dilemma. In Don’t Worry Darling, characters live in a simulated utopia without their consent. The Matrix questions whether we’d rather live a harsh truth or a pleasant illusion. WALL-E shows a future where humans become sedentary and disconnected while machines handle everything.
And Ready Player One presents a society where digital escape is preferable to the ruined real world. Each of these stories addresses different aspects of the experience machine (some voluntary, some compulsory), but all grapple with the tension between reality and artificial pleasure. Stories like these often serve as critiques of our real-life circumstances.
Generative AI could simulate not just environments and activities, but entire personalities, memories, and emotional responses. It could allow someone to fall in love, grieve, or grow—all within a fabricated world of their choosing. This wouldn’t just replicate life. It could replace it.
I’m not advocating for a complete ban on all technology, especially given the significant benefits it brings to society. However, we must ask not just whether we can build an experience machine, but whether we should, and who gets to decide what experiences are worth escaping to in the first place.
If the perfect life could be generated for you, custom-tailored to your dreams, needs, and emotions, would you plug in? And if so… would you ever come back?