ΑΙhub.org
 

“I am here to assist you today” – how we respond to chatbots

by
05 July 2021



share this:
Speech bubble | NLP

Photo by Volodymyr Hryshchenko on Unsplash

As online users we are increasingly exposed to chatbots as one form of AI-enabled media technologies. Stand-alone chatbots are often used for product or service recommendations, for example when online shopping or making financial or health-related decisions. What is the persuasive potential of these chatbots? Carolin Ischen studies how our perceptions of AI and our experience with chatbots influence our attitudes and behaviour.

Carolin Ischen is a PhD candidate at the Persuasive Communication group and the Digital Communication Methods Lab of the Communication Science department at the University of Amsterdam. She studies the persuasive potential of AI-driven media, specifically chatbots. “Chatbots reshape today’s media environment and move the communication science field from computer-mediated communication into human-machine communication. I therefore study technology as communicators and not only tools for communication”, tells Ischen.

In her research Ischen focuses on chatbots that recommend products or services, like a recipe or insurance, and how our perceptions of and experience with these non-human assistants influence our behaviour. “People tend to respond to computers in comparable ways as they do to humans. I look at how people enjoy the interaction with the chatbot and how they judge the content and brand information that is offered to them”, Ischen adds.

Carolin Ischen
Carolin Ischen

Enjoyment is the key mechanism

In a first experiment Ischen and colleagues examined the effects of interacting with a stand-alone chatbot compared to more traditional interactive websites. With the use of tailormade websites and virtual assistants they randomly asked a representative sample of the Dutch population to interact either with the chatbot or with the interactive website to receive a recommendation for a health insurance. Afterwards participants had to complete a survey to assess the experience: How human- or machine-like did the perceive the assistant? How much did they enjoy the interaction? Would they purchase the brand that was recommended?

Ischen et al found that enjoyment is the key mechanism explaining the positive effect of chatbots versus websites: the interaction with a stand-alone chatbot resulted in more enjoyable user experiences, which subsequently translated into higher persuasive outcomes. But contrary to expectations, perceived anthropomorphism (i.e., attribution of human-like characteristics), seemed not to be particularly relevant in this comparison. “The mere presentation of a chatbot as the source of communication as done in this study was likely not sufficient to increase human-likeness.”

Chatbots and privacy concerns

Ischen also looks at privacy concerns relating to chatbot-interactions. “While such concerns are widely studied in an online (website-) context, research in the context of chatbot-interaction is still lacking.” Ischen investigated the extent to which chatbots with human-like cues are attributed human-like characteristics and how this affects privacy concerns, information disclosure and recommendation adherence. She finds that people have fewer privacy concerns and disclose more information to chatbots when they perceive chatbots as more human-like, often leading to a follow up of the recommendation.

Ischen calls the study of chatbots in consumer interaction important in light of consumer awareness and empowerment. “As consumers we need to be aware that what is being communicated to us by chatbots is not neutral or bias free. Are we as consumers informed enough how this technology works and how to use it safely? What kind of regulations are needed for advertisement practices driven by AI?”

Further reading

“I am here to assist you today”: The role of entity, interactivity and experiential perceptions in chatbot persuasion, Ischen, C., Araujo, T. B., van Noort, G., Voorveld, H. A. M., & Smit, E. G. (2020).
Privacy concerns in chatbot interactions, Ischen, C., Araujo, T., Voorveld, H., van Noort, G., & Smit, E. (2020).




University of Amsterdam




            AIhub is supported by:


Related posts :



Keeping learning-based control safe by regulating distributional shift

We propose a new framework to reason about the safety of a learning-based controller with respect to its training distribution.
30 September 2022, by

Bipedal robot achieves Guinness World Record in 100 metres

Cassie the robot, developed at Oregon State University, records the fastest 100 metres by a bipedal robot.
29 September 2022, by

#IJCAI2022 distinguished paper – Plurality veto: A simple voting rule achieving optimal metric distortion

How can we create a voting system that best represents the preferences of the voters?
28 September 2022, by

AIhub monthly digest: September 2022 – environmental conservation, retrosynthesis, and RoboCup

Welcome to our monthly digest, where you can catch up with AI research, events and news from the month past.
27 September 2022, by

The Machine Ethics Podcast: Rights, trust and ethical choice with Ricardo Baeza-Yates

Host Ben Byford chats to Ricardo Baeza-Yates about responsible AI, the importance of AI governance, questioning people's intent to create AGI, and more.
26 September 2022, by

Recurrent model-free RL can be a strong baseline for many POMDPs

Considering an approach for dealing with realistic problems with noise and incomplete information.
23 September 2022, by





©2021 - Association for the Understanding of Artificial Intelligence


 












©2021 - ROBOTS Association