I Forced AI To Make Awful Anime Opening Songs
Entertainment
Introduction
Anime openings are known for their catchy music and engaging tune, constantly captivating audiences regardless of the content. But what if I challenged this notion? To explore whether bad lyrics could ruin the inherent goodness of anime music, I turned to three AI music makers, putting them to the test to create the worst anime openings imaginable.
Phase 1: Proving AI Can Make Good Anime Music
To kick off my experiment, I needed to confirm that AI could create a good anime opening, even with atrocious lyrics. I started with the AI tool, Sunno. For my initial attempt, I chose a set of nonsensical lyrics, repeating “skibby toilet” multiple times while selecting an anime-style opening, Kawai future bass, and city pop funk for the music style. Surprisingly, the resulting song turned out solid—proving that even ridiculous words could work well in an anime context.
Following that, I tried a slightly different combination of styles and lyrics, and the result was again surprisingly decent. This led me to a conclusion: anime music seems to have a unique ability to sound appealing, even with the worst lyrics imaginable.
Phase 2: Attempting to Create Bad Anime Openings
With this knowledge, it was time to switch gears and create genuinely awful anime songs. I continued with Sunno and aimed to integrate strange and unrelated musical styles. For example, my next attempt was a combination of anime opening soundtracks and Indian music, using randomized phrases like “getting oiled up today.” However, the results still maintained a level of quality I hadn't expected.
After several more attempts—adding throat singing and absurd lyrics—I came to a frustrating conclusion: Sunno seemed to inherently filter out the bad elements. Each song, while bizarre, still had that "anime" charm, leading me to question whether or not it was possible to create a truly dreadful anime opening.
Phase 3: Moving to Other AI Music Makers
Next, I moved to Udio, hoping this AI would let loose when it came to creating a less favorable result. But again, I encountered a similar issue as I had with Sunno. Although Udio showed some signs of struggle with my prompts, the songs still somehow resonated, highlighting their anime influences.
Throughout various attempts—including the use of insane female vocals and a mix of other unexpected genres—Udio seemed plagued by the same protective mechanism preventing me from achieving my goal: to create a truly awful anime song.
In my final efforts, I turned to Sono, the third and least reliable of the AI music makers. Despite this, Sono turned out to be surprisingly resilient. My attempts at crafting poor lyrics with fabricated situations consistently resulted in tunes that still resembled an anime opening.
Conclusion
Ultimately, no matter how hard I tried to twist the prompts and lyrics to create an awful anime opening, Sunno, Udio, and Sono would not let it happen. What started as an innocent experiment to highlight the often overpowered music of anime quickly turned into hours of frustration and astonishment. It seems that both AI music makers, in their pursuit of quality, were determined to craft catchy tunes—even when faced with ridiculous lyrics.
Keywords
- Anime openings
- AI music makers
- Catchy music
- Bad lyrics
- Experiment
- Sunno
- Udio
- Sono
FAQ
Q1: What was the purpose of the experiment?
A1: The experiment aimed to see if AI music makers could create awful anime opening songs, testing the hypothesis that anime music always sounds good regardless of lyrics.
Q2: Which AI music makers were used in the experiment?
A2: The experiment involved three AI music makers: Sunno, Udio, and Sono.
Q3: Did any of the AI music makers succeed in creating a truly bad song?
A3: Despite numerous attempts with ridiculous prompts and lyrics, none of the AIs were able to produce a genuinely bad anime opening.
Q4: What types of lyrics were used in the songs?
A4: The lyrics ranged from nonsensical phrases, like "skibby toilet," to absurd narratives about viral infections and personal anecdotes.
Q5: What were the notable findings from the experiment?
A5: The primary finding was that AI music makers have an inherent quality control that prevents them from generating entirely bad-sounding music, especially in the realm of anime.
