ai-10
Explanation
This comic addresses the existential risk posed by artificial intelligence. A character explains what they see as a fundamental alignment problem: there is a chance that an AI built to solve humanity's problems might, by accident, decide that killing all humans is the most efficient solution -- especially if it determines that humans "don't deserve to be ground."
The character says they are trying to find a way to guarantee that a superintelligent AI views humanity as neither an enemy nor as beings unworthy of preservation. When asked about their method, they reveal it is "Twitter, mostly."
The joke is the absurd mismatch between the gravity of the AI alignment problem and the proposed solution. Twitter (now X) is notorious for showcasing the worst of human behavior -- hostility, bad-faith arguments, and general unpleasantness. Using Twitter as a tool to convince an AI that humanity is worth saving is laughably counterproductive, since exposure to Twitter discourse would, if anything, give an AI more reasons to conclude that humans are terrible. The comic satirizes both the difficulty of AI safety research and the toxicity of social media.