4 Degrees of Anthropomorphism in Using AI

Oct 23, 2023

Summary: User research shows that people often treat an AI as another person at different degrees of fidelity. Anthropomorphism serves as a cognitive bridge, aiding users in understanding AI through human-like metaphors.

User testing has shown that people often interact with generative AI chatbots like ChatGPT at various degrees of anthropomorphism. Users treat the AI more or less as if it were another person. The 4 degrees classify how closely the user deals with AI as a person. You should read the full article, but here’s a summary of these 4 different behaviors:

First degree: Courtesy (FriendlyBot) includes niceties like “please” and “thank you” in prompts. While strictly unnecessary when talking to a machine, this brief nod across the abyss of cyberspace is a polite but shallow recognition of another entity’s simulated presence, devoid of deeper connection or purpose. The courtesy protocol is similar to referring to a dog by a human pronoun like “he” instead of referring to the dog as “it.” We acknowledge the dog as an independent entity worthy of some respect instead of being a mere object. Calling your dog “he” doesn’t mean you expect him to graduate from college.
Second degree: Reinforcement (KudosBot) is a more advanced form of interaction where users offer feedback to guide the AI’s behavior. For example, the user might tell the AI, “Good job!” Such virtual thumbs-ups are a step beyond mere politeness, offering a mechanism for shaping future interactions to better align with user expectations. The hope is that the AI will do more of what it’s being praised for, similar to what you would hope from a small child who’s praised when eating spinach.
Third degree: Roleplay (CosplayBot), where you tell the computer to pretend to be a specific type of professional or other specialized persona. For example, a prompt might be “You are a prize-winning pastry chef at a top hotel in Paris. Tell me how to bake the perfect croissant.” This enhances the quality, relevance, and utility of AI output, making the interaction more meaningful and purpose-driven, aligning the AI’s work with the user’s strategic objectives.
Fourth degree: Companionship (LoveBot) is the pinnacle of AI anthropomorphism, where the AI is not a tool but a companion. It offers emotional support, psychological benefits, and a form of engagement beyond mere functionality. While users still know deep down that they are interacting with a computer, they suspend disbelief and treat it as a friend or even a romantic interest, as shown by the virtual girlfriend/boyfriend services currently doing big business.

The 4 degrees of anthropomorphizing AI: FriendlyBot, KudosBot, CosplayBot, and LoveBot (artwork by Midjourney).

Anthropomorphism is a new type of metaphor in human–computer interaction. Traditional metaphors were mainly visual, from the traditional Macintosh desktop metaphor where deleting a file was visualized as a trashcan to Jony Ive’s skeuomorphic note-taking app visualized as a richly rendered notebook, complete with lines and a leather binding at the top.

Metaphors serve as cognitive bridges that help users understand new or complex systems by relating them to familiar experiences. You can transfer your knowledge of the reference domain (for example, a notebook) to learn how to function in the target domain (the note-taking app).

The four degrees of anthropomorphic metaphors similarly help users operate an AI by transferring their understanding of how to deal with fellow human beings.

As an exercise, I asked ChatGPT’s “Diagrams Show Me” plugin to visualize the 4 degrees of anthropomorphism. Here is its diagram. This is more informative than most infographics on the web, so kudos for remembering this angle that is often overlooked in superficial infographics. Visual attractiveness? The less said, the better. You still have a ways to go, my dear friend ChatGPT (exhibiting a bit of 2nd-degree behavior here, I guess). Accessibility? Maybe a C at best, mainly due to the small font size, which will cause problems for readers over 50.

Here is that same diagram formatted as an infographic. Which version do you prefer? Let me know in the comments.

More on AI UX

This article is part of a more extensive series I’m writing about the user experience of modern AI tools. Suggested reading order:

About the Author

Jakob Nielsen, Ph.D., is a usability pioneer with 40 years experience in UX and co-founded Nielsen Norman Group. He founded the discount usability movement for fast and cheap iterative design, including heuristic evaluation and the 10 usability heuristics. He formulated the eponymous Jakob’s Law of the Internet User Experience. Named “the king of usability” by Internet Magazine, “the guru of Web page usability" by The New York Times, and “the next best thing to a true time machine” by USA Today. Before starting NN/g, Dr. Nielsen was a Sun Microsystems Distinguished Engineer and a Member of Research Staff at Bell Communications Research, the branch of Bell Labs owned by the Regional Bell Operating Companies. He is the author of 8 books, including the best-selling Designing Web Usability: The Practice of Simplicity (published in 22 languages), Usability Engineering (26,238 citations in Google Scholar), and the pioneering Hypertext and Hypermedia. Dr. Nielsen holds 79 United States patents, mainly on making the Internet easier to use. He received the Lifetime Achievement Award for Human–Computer Interaction Practice from ACM SIGCHI.

· Follow Jakob on LinkedIn.

· Subscribe to Jakob’s newsletter to get the full text of new articles emailed to you as soon as they are published.

Jerome

Oct 23, 2023Liked by Jakob Nielsen

The inforgraphic is obviously more readable and for that reason more informative (?) but I personally appreciate the 'cyclic' viz that this particular diagram can offer: the visual idea that the user might be able to just skip and choose exactly the type of anthromorphism they need for that particular moment (at random or picking by momentum and decision).

So for example: don't be my companion when I don't need any (let me work, haha), but offer me clear visuals when I don't care tor read a listed infographic - even decorated with emojis ;-). I can think of a myriad of combinations of more or less motivated decisions; do we need to focus on one Degree only when crteating AI bots or can we think of an open architecture, responding with more accuracy and 'presence of mind' (such as a human may display at times)?

Expand full comment

Eli Schiff

Eli’s Shadow

Oct 24, 2023

> Jony Ive’s skeuomorphic note-taking app visualized as a richly rendered notebook, complete with lines and a leather binding at the top.

So you know, Jony Ive wasn't responsible for that leather or paper

2 replies by Jakob Nielsen and others

2 more comments...

Jakob Nielsen on UX

Discussion about this post