Skip navigation.

User Centered

Studying the design of everyday things

Just like me!, Techie Toys, part 3 of 3

, ,

There's something...creepy about machines that display human qualities. From the "we are sorry" used in Microsoft's "This program has encountered a problem" dialog, to the doll that says "dumb mama!" when she gets "mad," we don't like it when computers impersonate us.

The $40,000 word is "anthropomorphism" and it just means "things that are given human qualities." We do it ourselves, naturally, so there's no cause for alarm. If you've ever named your car or believed that your pet is "trying to comfort you," you've anthropomorphized them.

How does this relate to toys? Well, dolls and stuffed animals have always tried to be "more human." Before technology, they were made impressive by the quality of the materials used, clothes and accessories that look "just like" their owners full-sized counterparts. And with the myriad technologies available now (voice chips, sensors, motors, screens), it seems the more a toy can trick us into believing it's human, the better it sells. 20 questions, digital pets, anyone?

But when can we (or should we?) use the technique when developing user-focused systems? First of all, try not to do it by accident. It only takes a subtle shift in tone: "I need a valid telephone number" vs "Please enter a valid telephone number." (use the second one, btw)

Second of all, fit with your theme. A video game probably benefits from a "virtual guide," but why did most people hate Clippy? Wrong setting. People were trying to get stuff done, and this stupid character kept interrupting their work. Anthropomorphism rarely works in "regular" applications. Even Skype, a very fun-themed application, doesn't use "I" for itself (and only uses "we" on their website when it's clear they're referring to the people in their company).

And last of all, don't do it unless you can really pull it off. A five year old might be "fooled" by a talking doll (our niece actually got mad at Amanda when she said "dumb mama"). There isn't a system yet, however, that can pass the Turing Test.

Usability principle: Usable systems rarely benefit from anthropomorphism. Steer clear!

Also see:
Part 1 Amazing oversight: forcing functions, and Part 2 Some assembly required: Buyers as users.

The Pros and Cons of PieBand Aid Fixing

Comments

Eddie_Lopez 5. January 2007, 15:08

Ah... it's the Computer part of HCI. It's certainly true- I don't want my computer to act like my friend or confidant, because it's not and feels oogy when it tries to be.

So clearly my.opera.com's theme is "Primates working in IT that like to lose my requests/submissions on occasion"

Oh anthropomorphized server monkey...how I pine for thee.

nonrecursive 9. January 2007, 14:41

My recent brush with creepy computers - I submitted a story on Digg that had already been submitted, and the message was, "Cool! Someone else has already Dugg this!" or something like that. Why is a computer giving the message, "Cool!" ?

OmegaJunior 5. April 2007, 11:25

Why?

Why would we dislike a computer (application) or any other piece of technology for that matter, displaying antropomorphic qualities? What is the psychological benefit of this "ick" factor? Is it the same in all cultures? In all times?

Try robotics. Some organisations are building robots with faces that can mimic human expressions, as antropomorphs. Other organisations build robots without that quality. There seems to be a use for both. However, if the rule-of-thumb is that everyone feels icky about techo-antropomorphism, the antropomorph bot is doomed to fail from the get-go.

How to use Quote function:

  1. Select some text
  2. Click on the Quote link

Write a comment

Comment
(BBcode and HTML is turned off for anonymous user comments.)

If you can't read the words, press the small reload icon.


Smilies