How to spot a robot
Of course, when you’re just a regular person, it’s hard to tell if you’re online dealing with a robot or a human. As artificial intelligence and the tech that supports it advance, however, robots are becoming more realistic and easier to identify. Here’s how:
Use common sense. While it’s not always easy to spot, most bots are programmed to give the illusion that they are human. This is called chatbots, which means they use text-based messages to respond and interact with humans on websites and other sites. If a person seems robotic or cold, whatever you’re communicating probably is too — and it sure sounds like a bot.
Examine the details. Bots tend to focus more on small details in their conversation and will often miss things that might seem important to the average person. For example, they may forget to mention the weather, the time or even their own name. The lack of these details might lead you to suspect you’re dealing with a bot.
Experts note that chatbots are becoming more sophisticated, and that there are many different meanings for the word “bot.” Some interpretations include: “robotic,” “automatic” or “mechanical.” Some describe it as a type of chat software that can engage in a conversation and carry it out with little input from its creator.
“It’s a very intangible skill, and I think that’s why people are missing the mark,” said Rob Enderle, an analyst for the Enderle Group. “They don’t know what to look for.”
Why websites use CAPTCHAs
Captchas are letters and numbers used to filter out spam. To get access to a website, you must first complete a CAPTCHA and show that you’re not a robot. The very first reason these filters were invented was to prevent people from signing up for a site with automated software. Now that it’s easier to use more sophisticated software, it’s harder for bots to create accounts on a site and harass users. But most websites don’t want to keep spammers out of their site. They just want to make sure that users are human.
Chat rooms, forums and blogs allow users to post comments. These networks also need clear guidelines for posting content. CAPTCHAs prevent fake accounts from flooding a website with links to harmful or illegal content, or with links to advertisements that pay the spammer money. If a website has a lot of traffic that’s spread out over several servers, it can be hard to identify where the traffic comes from. CAPTCHAs help with this problem by identifying where users are coming from, and how many different sources they’re coming from.
How to solve a CAPTCHA
Paperclip, created by Flickr co-founder Stewart Butterfield, is a team collaboration tool which aims to improve productivity by simplifying the way that teams communicate online. The team, composed of a set of freelance developers and designers, has created a CAPTCHA, intended for use in mass posting.
A CAPTCHA is an acronym for “Completely Automated Public Turing test to tell Computers and Humans Apart.” It’s one of the most common methods for preventing spam on websites. It works by displaying a series of letters (usually images) that any human would have no trouble recognizing but robots would fail at correctly reading. The user then has to input their answer in order to continue.
A team of developers, designers and administrators would face several challenges in making Paperclip more user-friendly: they need to improve their productivity, they need to simplify their workflow, and they need to increase the amount of user participation.
In the paperclip team blog entry, “How to solve a CAPTCHA”, these goals are explained in detail.
Team members have identified four challenges that need to be solved in order to improve the team’s productivity:
1. Improve user experience and interface design so that CAPTCHAs are understandable without having to read the whole text of the comment. This is especially important in a team collaboration environment where the reader is not known, thus preventing your comments from being flagged as spam by bots.
The history of CAPTCHAs
The term CAPTCHA stands for “Completely Automated Public Turing test to tell Computers and Humans Apart”. The first CAPTCHA was invented in 1997 by Luis von Ahn, Manuel Blum, Nicholas J. Hopper and John Langford. The team was working on digitizing books at Carnegie Mellon University. They discovered that optical character recognition (OCR) software had trouble reading text in old books because the text was too old and degraded. To get around this problem, they programmed an OCR system to decode a user-submitted CAPTCHA image. The resulting string of letters and numbers was then matched against a specific text string on the page. This system of text-matching, hence the name “CAPTCHA”, was put in place to prevent automated programs from scanning online books. In 2000, Luis von Ahn and Manuel Blum started reCAPTCHA as a means of keeping spammers from creating fake sites that look like they belong to legitimate organizations by imitating the pages that steal personal information. To resolve this problem, they created a CAPTCHA that had the appearance of a human being reading the text. The results from using the program have been positive, and reCAPTCHA has been used by numerous well-known websites such as Google and Yahoo.
Interesting uses for CAPTCHAs
In addition to anti-spam technology and anti-dilution measures used in traditional stock markets, CAPTCHAs are used for a range of purposes, such as: ○ Solving captchas during online transactions ○ “Humans-reading” meeting minutes to determine if the meeting was actually attended by humans ○ Anti-intrusion systems for commercial buildings that replace the need for security guards (see also OCR) ○ Online voting to prevent people from voting multiple times ○ Comparing solved captchas with a database to determine if certain words are in a language. For example, the site for the dictionary of the Dutch language has a small “captcha” box, which asks whether or not you considered the definition of the previous word (such as “sleep” or “sneeze”) when choosing your answer to a given word. ○ Employed by Google and Yahoo! as part of their anti-spam measures