CPU-Opponents & Me#

Probably two of my favorite titles growing up, I often played against CPU-opponents in Mario Kart and Super Smash Bros. To most, CPUs are simply filler to populate an otherwise empty racetrack or a reluctant substitute for when our friends don’t want to play against us anymore. That hasn’t stopped Nintendo from adding little touches that spice up gameplay. In Mario Kart’s case, there’s an underlying rival system that dictates what two CPU racers will regularly cause you trouble throughout the cup. It’s easy to grow frustrated with the game when random racers are causing you trouble. However, when those incoming red shells are coming from a consistent source, you’re bound to personally relish ripping first place back from the CPU who dared to target you at all. Nintendo wants you to invest time into the game, and what better way than to create artificial rivals for you to beef it out with on the track?

Rivals System: Source

Speaking of frustration and beefing, my other favorite title, Smash Bros, also has CPU opponents with adjustable difficulty. The difference in difficulty is actually how likely the CPU is to commit or follow up on a move. Some Smash players have actually criticized Smash CPUs because their reaction times may occasionally be unrealistic for players and thus make them unfit as a serious competitive training method. I’ve personally never run into that issue myself. What bothers me, though, is why they wave dash after brutally spiking me offstage. Or why emote onstage after performing a nasty 0 to 80 combo while I’m still mentally recovering. Emoting has no functional purpose besides playing mental games against your opponent and Nintendo’s effort to simulate that, I’m afraid it’s working. It happens so infrequently that, when it does happen, you can’t help but lean forward and take the match seriously. You wouldn’t lose against a computer opponent, would you?

I hope these two examples help illustrate Nintendo’s efforts to personally invest their audience in their titles by appealing to their competitiveness. I’ve been using the term CPU this whole time, but it’s often used interchangeably with NPC and AI in the gaming space. I wouldn’t be surprised if CPUs and NPCs are gradually phased out of the gaming lexicon as AI continues to play a bigger part in gaming. Regardless, will AI incorporate the same quirky little features meant to rile gamers into more? It probably will and with varied success just like the current methods. AI, like the CPUs in this case, are executing algorithms to appeal to the emotional aspect of players. Frustration, competitiveness, sportsmanship, and other emotions that drive people to keep playing. AI doesn’t reciprocate these emotions though, and it’s this emotionless quality that I’d like to explore when pondering the question: Can AI create art?

What’s AI?#

But first, let’s lay some groundwork. What exactly is AI? Artificial Intelligence is a field of study focused on creating machines and services that can complete complex tasks typically requiring human reasoning. There are plenty of activities that, at first glance, seem like things only humans can complete. For instance, it was originally believed that a computer could not defeat a chess player. Although personally a novice player, I recognize that chess is an incredibly nuanced game that goes beyond its simple rules. You need to be adaptive and flexible. You need to be able to read your opponent and determine the best course of actions five, maybe even ten, steps ahead. It simply was unimaginable that an early computer could ever beat a chess player. Yet, the first domino fell in 1956 when the early computer MANIAC managed to defeat a novice chess player in 1956. It took another 41 years for IBM’s supercomputer Deep Blue to defeat World Chess Champion Garry Kasparov. It wasn’t just chess though, somewhat recently DeepMind Technologies’ created the program AlphaGo. At a time when computers were still struggling to compete against Go players, AlphaGo would eventually advance the competitive ladder and have the honor to compete against South Korea’s Go champion Lee Sedol in a 5-match game. Most viewers believed that AlphaGo could not possibly defeat the legendary Mr. Sedol. Unfortunately, things didn’t quite result the way people imagined it. AlphaGo would claim three consecutive matches before Mr. Sedol claimed the fourth. There was no fifth match, Mr. Sedol resigned the final game and said “I, Lee Se-dol, lost, but mankind did not”. It’s frightening to admit, but most tasks that have defined rules with clear terminal states will be completed by AI programs and will surpass mankind in capability.

However, the one realm that AI continues to struggle in is replicating the emotional. Humans are rational beings, but we also crave social connections. We’ve evolved to seek others’ help because being part of a community significantly impacts your likelihood of survival. Although some may think differently today, it’s undeniable that friends provide emotional fulfillment that is critical to being healthy. For all the courtesy and professionalism of ChatGPT, Gemini, Llama, and any other AI assistant, they cannot create artistically. Many AI models are unethically, and sometimes illegally, trained by scrapping the Internet for vast amounts of material: social media posts, books, songs, poems, artwork, conversations. If it was publicly available on the Internet, then there’s a good chance it’s been gobbled up by a generative AI model for training. The AI model takes all of this input and, when prompted by a user, creates a response by analyzing the user’s query and calculating their intent to provide the most relevant and usually factual information. It’s excellent at answering basic queries and performing simple tasks, but its artistic creations feel too robotic. Unoriginal would be another way of describing it. It feels like something else I’ve read someplace else. And of course something isn’t made from nothing, many artists cite numerous past works that inspired their own. However, generative AI models are not pulling from the past to create their own original works like human artists are. As cliché as it sounds, each human artist pours a little bit of themselves into their work. The emotional realm that I mentioned earlier, that is added to their work to make it their own. AI cannot add anything itself. It doesn’t create the work for itself, it creates it because a user prompted it to.

Twitter & AI Art Question#

Frankly, the primary drive behind this opinion piece was the official White House’s Twitter account posting an inappropriate “Ghiblified” rendition of a deportation.

White House Twitter Account: Source

I found this publication quite disturbing because it brazenly presents the serious topic of deportation and drug trafficking in a non-serious manner. I’m uncertain as to the reason behind the Ghibli-inspired image: was it a personal decision by Elon or one of his staff to troll the libs? An effort to appeal to a broad age demographic? Whatever the case, it speaks to an administration’s noncommittal to the severity of its own actions. Despite how you may feel about immigration, this feels wildly inappropriate.

The news also created a wave of misinformation. A 2016 interview has begun to re-circulate where Miyazaki was asked his thoughts about AI art, to which he responded, “an insult to life itself”. Whether Miyazaki continues to hold this view is another matter, but it’s troubling that misinformation quickly spreads and even more deeply disturbing that companies are so quick to potentially violate artists’ copyright for the sake of business growth.

Speaking primarily as someone who semi-frequently surfs the cesspool that is Twitter, the people who advocate for AI art typically have a business mindset with their qualification. AI art accounts, from personal observation, advertise their services by creating AI renditions of people’s art without their consent. Clearly disrespectful, it’s made more infuriating by offending individuals playing victim and refusing to understand why people are upset and mocking them for their idiotic choice.

Fixed the shading in this one and applied Studio Ghibli style, I hope you like it!! pic.twitter.com/DGmtadO4NS

— Young One Dev (@YoungOneDev) April 8, 2025

Others discredit artists by claiming that AI art is art by virtue of the fact that people are willing to pay for it. That’s simply a grift. If you really want to create an AI generated image, then visit the numerous free services available online with a nice Adblock pre-installed to avoid the nasty ads. You don’t need to pay someone to enter a prompt that would’ve taken users seconds to do themselves. If you want something that a generative AI model can’t produce, commission an artist. If you don’t like their prices, then shop around. This odd appeal of monetary value justifying something classification as art feels strangely compensative in nature. AI “artists” should own their beliefs in some non-superficial way. If they genuinely believe it’s art, then they own it instead of putting others down. It’s an enriching dialogue to be held here about the ethics of generative AI image models, but it’s being limited by business grifters and others unwilling to engage with the topic.

At the end of the day, AI generated images will not go away. They will continue to improve and potentially reach a point where even the trained artist eye has difficulty distinguishing between AI generated images and art. If you want to hear a good song about this, there’s exocity’s THINGS WILL GET MUCH WORSE FROM HERE. Just felt like plugging something at the end.

Citations

1. “The Computer Beats a Chess Master.” Byte Magazine, Dec. 1978. Internet Archive, https://archive.org/details/byte-magazine-1978-12/page/n85/mode/2up?view=theater.
2. IBM’s Deep Blue Defeats World Chess Champion Garry Kasparov (1997). CNN, 11 May 2022, www.cnn.com/videos/business/2022/05/11/ibm-deep-blue-computer-beats-garry-kasparov-chess-champion-1997-vault-jg-orig.cnn.
3. AlphaGo. Directed by Greg Kohs, performances by Demis Hassabis, Lee Sedol, and Fan Hui, DeepMind, 2017. YouTube, uploaded by DeepMind, 28 May 2020 https://youtu.be/WXuK6gekU1Y?si=ljvth0Y8UfjrE36z.
4. “Insult To Life Itself”: Ghibli Founder Hayao Miyazaki On AI-Generated Art. NDTV, 12 Apr. 2025, www.ndtv.com/world-news/quot-i-would-never-incorporate-this-quot-what-studio-ghibli-039-s-hayao-miyazaki-once-said-about-ai-animation-8021037. :::