2-day virtual workshop
June 6-7 2024, 12:00pm-4pm PT
Real-world artificial Intelligence can seem overwhelming. It’s a complicated thing and when you try to look into it, it’s filled with intimidating computational jargon. For designers, though, the basics of AI can be made pretty simple. By understanding the basics of data, models, algorithms, and machine learning, you can make sense of almost any AI, separate out what is meaningful for design, and continue with your user-centered practice, mostly as before. (And yes, we’ll cover generative AI.) In this workshop we’ll go over these basics, talk about the interactions germane to narrow AI, and get some hands-on practice in teams.
Target Audience
This workshop is meant mostly for mid-level designers already familiar with the basic practice of user-centered design, and who want a practical, actionable understanding of what AI is and does for users.
NOTE: Attendees will not learn new and nifty AI tools to use in their practice. (Like, say, Midjourney or chatGPT/Dall·E.) If there’s time in the Q&A and interest from the attendees at the end, we can discuss and share some, but this workshop is about designing for systems that include insights and recommendations from AI. (There has been confusion about this in the past, so good to emphasize.)
Pre-requisites
Users should have a working knowledge of user-centered design, especially from an interaction design perspective.
Take-aways
- Attendees will be able to name the 4 fundamental components of modern AI and know how they fit together to deliver user value
- Attendees will have a working understanding of the 4(ish) core algorithms underneath modern AI
- Attendees will understand how the 29 core interactions are derived from the core algorithms
- Attendees will be more confident in coming up with design ideas that involve AI and design for them with users still at the center of their practice
- Attendees will be able to contextualize machine learning in terms of user needs
Agenda
- Pre-start: Inviting people to introduce themselves and their interest in AI on a shared board (like Miro/Mural)
- General intro: Breaking AI into manageable (and learnable) parts, in a rough chronological order.
- Data
- Machine Learning (making models, yes, it’s coming for users)
- Models
- Algorithms
- Introducing the 29 core interactions
- Machine Learning (feedback)
- Algorithms: Introducing the core categories
- Clusterers
- Classifiers
- Regressors
- Generators
- Optimizers
- Introducing a toy problem
- Exercise: Picking and designing for one algorithm-related interaction
[BREAK]
- Going deeper into Models
- Maps: A very instructive metaphor
- An FAQ
- Model-related interactions
- Exercise: Picking and designing one model-related interaction
- A designers’ view on Machine Learning (feedback)
- Introducing modes of agency
- Manual (this almost doesn’t exist anymore)
- Assistive
- Agentive
- Automated (this almost doesn’t exist)
- Exercise: Redesigning one of the prior designs in the context of agency
- Nano-primer: Keep your eyes open for deskilling and overreliance
- Wrap-up and call-to-action: You got this
- Q&A
