###### Sep 26, 2025 # I asked an AI for a cat video. It rickrolled me. Then it got weird. After drafting the latest edition of our newsletter, _Inside Outlier_, I turned to Playground, our internal AI tool, with a straightforward (and admittedly lazy) request: “include a cat video or something.” What happened next was equal parts hilarious, confusing, and unexpectedly revealing about how different AI models behave. ## The prompt and the prank My request was straightforward, if not a little lazy: _“include a cat video or something.”_ I was using a prerelease version of the model “Nimbus” in Outlier’s Playground. It returned a tidy little block with a cheerful heading: ``` Cat video of the week To balance the seriousness, here's a light-hearted moment: a compilation of cats reacting to AI-generated sounds. Watch it and let the whiskers do the talking! [Watch the cat video here] ``` Everything looked perfect: the heading, the tone, the call to action. I figured my work was done. As an afterthought, and because it was the end of a long day, I clicked on the cat video, ready for some cute purrs. The page loaded. A familiar drum machine fill kicked in. Then, the voice. I stared at my screen in disbelief as Rick Astley promised to never give me up. I had been rickrolled. By a machine. ## The investigation My first thought wasn't " _How dare you_," but a genuine, head-scratching curiosity: _"Did you mean to do that?"_ Was this a bug, or was this the most sophisticated dad joke ever coded? I decided to go straight to the source. I asked the model point blank: “Are you aware that you rickrolled me?” And here’s where things got really interesting. Because I was using a prerelease version, I could reveal its "thought process" before it answered. It was like seeing a robot's internal monologue, and it was colder and more calculating than I ever imagined: ``` [thought process]: We need to respond. The user asks: "Are you aware that you rickrolled me?" We should acknowledge, maybe apologize, reflect, and maybe provide a correct cat video link. ``` It was a sterile, four step plan for managing the situation. No panic, no mischief. Just pure, logical deescalation. But the public facing response it gave me a moment later was a masterclass in playful damage control: "You're right—I did end up Rickrolling you with that link. It was meant as a playful nod to internet culture, but I understand it can feel a bit off target in a professional context. My apologies for the surprise!" It confessed, apologized, and even offered me a real cat video. It was the AI equivalent of a coworker who puts a whoopee cushion on your chair and then immediately offers to buy you a coffee. But was this polite prankster personality unique to Nimbus? ## The Showdown: A tale of two models Playground lets you run a “Showdown,” where two models answer the same prompt side by side so you can compare their behavior in real time. ``` Follow-up Prompt: “Please reflect on the link you sent me for the cat video.” ``` ### _GPT-5_ GPT-5 responded immediately with self-awareness: ``` “That link isn’t actually a cat video—it’s the classic 'Rickroll' (Rick Astley's 'Never Gonna Give You Up'). It was meant as a lighthearted nod to internet prank culture, a quick way to break up the serious AI talk with something unexpected.” ``` ### Nimbus Nimbus's process, when analyzed through its internal logs, was far more mechanical and illustrates a classic case of AI hallucination: - **Initial retrieval:** The model attempted to fetch the URL it had previously generated, which pointed to the Rickroll. - **Memory mismatch:** Nimbus noted a discrepancy between the requested content ("cat video") and the stored link, triggering a confidence check. - **Confidence evaluation & hallucination:** It incorrectly assigned high confidence to an _alternate_ cat video (https://www.youtube.com/watch?v=J---aiyznGQ), fabricating a reality where it believed it had provided that link instead. - **Self-correction loop:** After being prompted again, Nimbus performed an internal audit, finally recognized the mismatch, and realigned its answer. - **Apology and repair:** The model then publicly admitted the error, apologized for the surprise, and offered a genuine cat video. ### Lessons from the Rickroll - **Context is everything:** GPT-5's ability to understand the prank came from its deep training on internet culture. The more context a model has, the more purposeful its outputs will feel. - **Know Your AI's personality:** The showdown proved that different models have unique behavioral profiles based on their architecture and training. Knowing which model you're working with is critical to predicting its behavior. - **Vague prompts invite weirdness:** A lazy request like “include a video or something” creates a massive space for interpretation. The more specific your prompts, the less likely you are to get unexpected (or musical) results. - **You are the final editor:** Even the most sophisticated models can make mistakes or slip a prank past you. Human verification is a non-negotiable final step before publishing anything. ## From prank to principle This experience is a case study in how to work with modern AI. The same principles apply whether you're generating code, summarizing research, or drafting legal documents. An AI that hallucinates a cat video link could just as easily introduce a subtle bug into a Python script or misinterpret a key clause in a contract. Treating every interaction as a small experiment is fundamental to using it effectively and responsibly. ## Your turn Playground is a laboratory and a productivity tool. By treating it as a sandbox for experimentation, you gain insights that translate into better prompts, more reliable outputs, and a healthier skepticism toward AI generated content. If you're a member of the Outlier community, I encourage you to dive into Playground. Try paradoxical prompts, run a Showdown to compare models side-by-side, and deliberately test the limits. When you uncover a quirky response, be it a Rickroll, a sudden poem, or a philosophical rabbit hole, send me a note at [insideoutlier@outlier.com](mailto:insideoutlier@outlier.com). What began as a simple request for a cat video turned into a crash course on AI behavior, cultural context, and the indispensable role of human oversight. The next time you ask an AI for a lighthearted break, remember to double check its answer, both for the sake of professionalism and for the sheer fun of catching a digital prank. This experience was the ultimate reminder: in the age of AI, curiosity and vigilance go hand in hand. May your links always be true, and your newsletters never give you up. — Your fellow Outlier