Three Australian outlets publish opinion articles that challenge public impressions of artificial intelligence (AI). The writers argue that AI systems can appear influential and widely adopted, yet still produce results they describe as limited, error-prone, or less capable than expected. Across the pieces, the central question is whether people should be “impressed” by AI when it sometimes performs poorly or demonstrates misunderstanding comparable to that of an untrained person. The articles do not report new technical findings; instead, they focus on observed gaps between AI’s perceived potential and its real-world performance. By framing their critiques as questions about public perception, the authors urge readers to look beyond hype and evaluate AI capabilities critically, especially when AI output affects decisions or information consumers rely on. The shared theme is skepticism about how AI capability is portrayed versus how it behaves in practice.