Blink if you’re human
The article explores the famous "blink rate" claim from Malcolm Gladwell's book *Blink*, which suggests that experts can make snap judgments as accurate as deliberative ones. It examines the scientific evidence behind this idea, discussing studies on thin-slicing, unconscious decision-making, and whether rapid cognition truly rivals careful analysis.
Background
- The article discusses a technique where someone claims to be able to tell if another person is a human or a bot during a brief video call by asking them to "blink" on command. The idea is that bots/AI cannot produce a natural, spontaneous blink in response.
- This ties into ongoing tech discourse around AI detection, CAPTCHAs, and the growing inability to distinguish humans from bots online. Reverse Turing tests are now a common anxiety.
- The piece reflects a broader cultural moment: as AI-generated video and voice become convincing, people are searching for simple, reliable signals of human presence — and finding that even basic biological cues (like blinking) are hard to fake naturally.
- dynomight is a popular independent tech/culture writer known for clear, thoughtful essays on AI, statistics, and human behavior.