AI Slop is Your Choice: Don’t Outsource Your Judgment

Welcome to Season 2 of CircleBudapest! So yup, AI. We use it a lot. But the more it produces, the more everything starts to look the same. Same brands. Same posters. Same copy. Same ideas, just faster. People universally call it: AI slop. And we don’t think AI is really the problem. Slop happens when making something becomes so easy that nobody stops to ask if it’s actually any good. That’s what this evening is about. AI can do more of the work. Sure. So what are we still responsible for? The Agenda & What You'll Learn 17:40: Doors open, drinks, people 18:10: Welcome to Season 2 18:15: AI-Assisted, Not AI-Generated (by Nándor Fehér) First, the mindset shift. No, this isn't a “stop using AI and go back to Figma” talk. Quite the opposite. We've built parts of our upcoming CircleCampus project with AI tools. These tools can do a ridiculous amount of work for us. But someone still has to decide what the thing should be, notice when it feels wrong, kill the generic ideas, and refuse to ship the first plausible answer. That's the human part. We'll get into AI slop, taste, visuals, copy, thinking with AI, and why obvious AI work often says more about the person who approved it than the model that generated it. We will also share a base prompt we recommend to make chat interfaces like ChatGPT less agreeable and way more useful when he actually wants his thinking challenged. 18:30: Stop Chatting With AI and Give It a Job (by Matt Malishev) Okay, so we've kept our judgment. Now let's make the machine useful. Matt works with AI systems that actually do things, not just sit in a chat window waiting for another prompt. What is worth automating? What tools are actually useful? What can be connected together? And what needs to happen before you'd comfortably let an AI system run without watching it every five minutes? Less “look what AI can do.” and more “here's where this actually saves you work.” 18:45: When the AI Does It, Who's Responsible? (by Bence Gajdácsi) And then there's the slightly less exciting part, the legal side. Once AI starts touching customer data, making decisions, generating work for clients, or acting inside your business, “the AI did it” stops being a particularly useful explanation. Bence will look at the practical legal side: what you can put into these systems, what you should probably think twice about, where regulation is heading, and who actually carries the responsibility when something goes wrong. Worry not, this will not be a lecture on EU legislation. It's the the questions you should ask before connecting everything to everything. 19:00 onwards: drinks, conversations, etc. Who Will Be In The Room? If you use AI already, but don’t want your work to start feeling generic, this is for you. If you’re figuring out what to automate, what to keep human, and where “good enough” starts becoming a problem, you’ll fit right in. Founders, freelancers, creatives, operators, developers. You definitely don’t need to be an AI expert. You just need to care whether the work is actually good. Location We'll be hosting at the fantastic KLUSTER Coworking space. The Legal Bit We'll be taking photos and videos during the evening. By attending, you consent to being in them. See you on the 18th. Bence & Matt & Nándi