Skip to main content

Telling Computers and Humans Apart through Error

Captchas (Completely Automated Public Turing Test to Tell Computers and Humans Apart) are a type of security measure designed to differentiate between human users and bots or automated programs. They are commonly used to protect websites and online services from spam or malicious activity by requiring users to prove that they are human before they can access certain features or complete certain actions.

Captchas typically consist of a challenge-response test that is easy for humans to solve but difficult for bots to complete. Examples of captchas include images with distorted text that users must type out correctly or audio recordings of words that users must transcribe.

The original captcha was created by researchers at Carnegie Mellon University as a way to protect the online ticketing system for the Pittsburgh Steelers football team from bots that were trying to purchase large numbers of tickets.

To Err is Human

Despite their widespread use, captchas have several limitations. One major limitation is that they can be difficult for humans to solve, especially for those with visual impairments or other disabilities.

Additionally, as bots become more advanced, they are increasingly able to solve captchas, making them less effective at distinguishing between humans and bots. Finally, captchas can be bypassed using a technique called "pharming," in which low-paid workers are paid to solve captchas for use in bot networks.

As technology continues to advance, it is likely that bots and artificial intelligence will become better than humans at solving captchas and other bot-detection tests. 

This will inevitably require adopting new methods for distinguishing humans from bots. One potential approach is to look for errors that humans are more prone to make than computers. Identifying such human-specific errors in tests or content is a possible way to determine whether a test-taker was human or if a piece of content was likely written by a human rather than an AI.

If a piece of content has errors in spelling or grammar, it is more likely to have been created by a human, as AI systems are generally able to produce error-free content. This can lead to a "celebration of error," as people begin to accept and appreciate the presence of errors in content as a validation of their human origins.

The idea of celebrating errors as a way to identify human authorship may seem absurd at first, but it highlights the fact that humans are fallible and makes mistakes, while AI and bots do not (at least not in the same way). This could be seen as a way to embrace human frailty and celebrate the unique qualities that make us human.

However, it is important to remember that while errors may indicate that a piece of content was created by a human, they do not necessarily make the content more valuable or worthwhile. Ultimately, the value of content should be determined by its quality and relevance rather than whether it was created by a human or an AI.