UNDERTONE
  • ABOUT US
  • UNDERTONE.AI
  • BLOG
  • CONTACT
  • ABOUT US
  • UNDERTONE.AI
  • BLOG
  • CONTACT
Search

Ai-first language design

7/9/2017

0 Comments

 
Picture
*Our article was originally published in Protytpr's Medium publication.

AI-first language design
​

“​Understanding humans is essential to the design and experience of a technology.”

​As we enter this AI-first realm, interdisciplinarity will be at the heart of design. It’s no longer viable to think that a single designer should be the sole “DRI” (Directly Responsible Individual) when it comes to crafting user experiences. As AI quietly creeps into our lives, voice user interfaces and conversational interfaces will play a major role in shifting our focus from the screen to the scene. In other words, mobile-first will become AI-first, and NLP (Natural Language Processing), NLU (Natural Language Understanding), and ASR (Automatic Speech Recognition) will shape contextual personalization as we know it. As Element AI’s Experience Designer, Masha Krol notes, “As AI-first design becomes more prevalent, we hypothesize that as designers, we’ll need to foster a new kind of relationship with technology, based increasingly on collaboration.”

Perhaps this may explain why my work as a Content Strategist and UX Writer has dramatically changed over the years. Long gone are the days of “filling in” Lorem Ipsum-filled mockups and dealing with messy version control files. (Shout out to Abstract! 🙌🏼) Language is integral to the UX Design process, and as such, it should be considered right from the start.Language is design. Therefore, writers and language specialists should be working alongside designers to build seamlessly intuitive experiences.

With job titles and roles ranging from UX Writing, Content Strategy, Conversational Optimization, Copywriting, and more, it’s no wonder the lines get blurred. Title aside, these key team members all play an important role in shaping the user experience and setting the communicative tone. At the end of the day, we’re all striving to create a communicatively competent experience that conveys trust, builds lasting relationships, and brings enjoyable value to users.

“With clear, consistent, personalized, and contextual messaging, language design can be used to strategically enrich user experiences in a way that transforms brands, helps businesses grow, inspires change, and makes the world a better place.”Enter: Language Design, the art of designing language for the end-to-end user experience.

Language Design, in all of its forms, is at the heart of it all. It affects product design, communications, user research, marketing, branding, social media, localization, and AI-powered conversational spaces. For such an little-known field, that’s a BIG list!

As we make this AI-first shift, it’s critical that we understand the fundamental challenges and key issues involved in tackling this intricate process. Sure, it may seem like there’s a world of endless opportunities out there, and to some extent that’s true. However, as conversational and voice interfaces become more prominent, there are many factors to consider if we want to ensure that we’re designing ethically responsible conversational environments. The key to designing such environments was masterfully summed up by Masha Krol, Experience Designer at Element AI. She created a “Guide to Symbiotic Human-AI Experiences” for AI-first design.
​
Picture
Masha Krol’s awesome AI-first Design “Guide to Symbiotic Humain-AI Experiences” — Element AI


​AI-first Language Design Considerations
​

Although her guide isn’t language-specific, there’s a clear correlation that addresses the linguistic aspect of AI-first design. Inspired by Masha’s initiative, our team at Undertone.io decided to develop our own version of language design-specific principles to bring clarity to the linguistic process involved in AI-first design.
​

1. Design for Discussion and Learning
  • Use language to render clear, consistent, and communicatively competent (Hymes, 1964) interactions.
  • Strive to design a cooperative learning (Deutsch, 1949) experience that’s engaging and beneficial for both parties.
  • Design for productive two-way conversations. Contextual analysis will allow the AI to scaffold (Bruner, 1957) within its Zone of Proximal Development (Vygotsky, 1978) as it activates its prior knowledge to bridge gaps and reach a more complex metacognitive understanding.This will allow the AI to develop its taxonomical understanding of objectives (Bloom, 1956).
  • AI isn’t perfect. Ask for feedback when it’s appropriate and encourage users to participate actively in helping the AI learn.​

​2. Design for Intelligence
  • Not only should your language render clear, consistent, and communicatively competent interactions, but linguistic choices should be thoroughly considered. As a language design specialist, you should be ready to explain your linguistic choices and demonstrate how they provide value to the user, and the AI.
  • Language should be as universally accessible as possible, yet contextually personalized, too.
  • Use language to build trust by connecting users to a personalized experience in meaningful context.

3. Design for Evolution
  • Design language with adaptation and change in mind.
  • Not only do brands evolve, but so does language, as well as your AI’s intelligence level as it learns from its users’ preferences and contextual environments. Language should be adaptative.
  • As technology evolves we will eventually reach human-like natural language understanding, thus allowing AI to potentially develop multilingual communicative competences.

​4. Design for Comfort and Delight
  • Design language in a way that exceeds your users’ expectations.
  • Design language for accessibility, differentiation, and cultural sensitivity.
  • Design with Gardner’s Multiple Intelligences Thoery (1983) in mind. We all have different learning styles, which means multimodal approaches should be implemented to facilitate the progression of learning.
  • Understanding your users, the context, and your users in context will allow you to craft meaningful interactions that build trust.
  • Understand the user’s emotional and rational behaviour in context and its effect on the overall user experience.
  • Use language to support users and make the most uncomfortable of situations comfortable. Acknowledge change, be gentle, and explain these changes to users.
  • Be honest and transparent. Open communication is the key to lasting relationships. Use data to explain behavior.
  • Be helpful, not intrusive. Try to limit one-off interactions and design with purpose and the end goal in mind.
  • Be proactive. Anticipate your users’ needs and use language to offer useful suggestions, make interesting recommendations, and request clarification to improve the overall user experience.
  • Voice and tone should be environmentally and contextually adaptive.
  • Use language to “humanize” conversations and bring warmth to interactions.
  • Allow users to feel in control by respecting sensitive boundaries. Understand when to help and when to hold back. As Masha Krol suggests, “include an explicit shut up button” and allow users to decide when they’ve had enough.

​5. Design for Purpose
  • Language design should be used to bring value to users in helping them achieve a specific end goal.
  • Keep it simple! Analyze behavioral patterns and context to reduce the amount of menial tasks your user has to execute.
Although this may appear to be an extensive list, please keep in mind that language design is continuously evolving, as is AI. These principles should be used to guide your language design process as you push forward with AI-first design. Again, these principles were originally written by Masha Krol, Experience Designer at Element AI. Our team has tweaked them to address the linguistic aspect of AI-first language design.
Thanks a million for reading my first AI-first article! I do hope that you’ll send feedback my way as I am always eager to read up on your approach to Language Design in an AI-first context. Also, feel free to share the article with your colleagues, and please get in touch via @Samb418 or hello@undertone.ioif you’d like to discuss Language Design further.

A huge thanks goes out to Masha Krol for meeting with us at Startupfest/AIFest 2017, and sharing her AI-first design “Guide to Symbiotic Human-AI Experiences.” Mad props also goes out to the entire Element AI team. Your work and initiatives continue to inspire us! 🙌🏼
​
Stay tuned for more Language Design articles! We’ll be sharing more about our Undertone.io Language Design Lab experiences, and AI-first design work processes in the coming weeks. Talk soon! 🤓 -Sam
Picture

Sources:
Bloom, B. S. (1994). “Reflections on the development and use of the taxonomy”. In Rehage, Kenneth J.; Anderson, Lorin W.; Sosniak, Lauren A. Bloom’s taxonomy: A forty-year retrospective. Yearbook of the National Society for the Study of Education. 93. Chicago: National Society for the Study of Education. ISSN 1744–7984.
Brown, H., & Ciuffetelli, D.C. (Eds.). (2009). Foundational methods: Understanding teaching and learning, p. 508. Toronto: Pearson Education.
Bruner, J. S. (1975/76). “From communication to language: A psychological perspective”. Cognition, 3, 255–287.
Bruner, J. S. (1978). “On prelinguistic prerequisites of speech”. In R. N. Campbell and P. T. Smith, (eds.), Recent Advances in the Psychology of Language (Vol. 4a. pp. 194–214). New York: Plenum Press.
Deutsch, Morton (1949). “A theory of cooperation and competition”. Human Relations. 2: 129–152. doi:10.1177/001872674900200204.
Deutsch, Morton (1962). Cooperation and trust: Some theoretical notes. Lincoln: University of Nebraska Press. pp. 275–319.
Gardner, Howard (1983), Frames of Mind: The Theory of Multiple Intelligences, Basic Books, ISBN 0133306143.
Gardner, Howard (2006), Multiple Intelligences: New Horizons in Theory and Practice, Basic Books, ISBN 978–0465047680.
Gardner, Howard (1993), Multiple Intelligences: The Theory in Practice, Basic Books, ISBN 046501822X
Gilles, R.M., & Adrian, F. (2003). Cooperative Learning: The social and intellectual Outcomes of Learning in Groups. London: Farmer Press.
Hymes, Dell (1964), “Toward ethnographies of communication”, American Anthropologist, 66 (6 part 2): 1–34, ISSN 0002–7294.
Hymes, Dell (1966). “Two types of linguistic relativity”. In Bright, W. Sociolinguistics. The Hague: Mouton. pp. 114–158. OCLC 2164408.
Krol, Masha (2017). AI-First Design Guide to Symbiotic Human-AI Experiences. Element AI.
Leung, Constant (2005). “Convivial communication: recontextualizing communicative competence”. International Journal of Applied Linguistics. 15(2): 119–144. ISSN 0802–6106. doi:10.1111/j.1473–4192.2005.00084.x.
Ross, J.,& Smythe, E. (1995). Differentiating cooperative learning to meet the needs of gifted learners: A case for transformational leadership. Journal for the Education of the Gifted, 19, 63–82.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.
Wood, D. J.; Bruner, J. S.; Ross, G. (1976). “The role of tutoring in problem solving” (PDF). Journal of Child Psychiatry and Psychology. 17(2): 89–100. doi:10.1111/j.1469–7610.1976.tb00381.x.
Wood, D., & Middleton, D. (1975). A study of assisted problem-solving. British Journal of Psychology, 66(2), 181−191
0 Comments



Leave a Reply.

WE  DESIGN  LANGUAGE  ❤️

Powered by Flip Script Media 
  • ABOUT US
  • UNDERTONE.AI
  • BLOG
  • CONTACT