SXSW 2024

AI by Design

Description:

Most AI projects fail. Some fail quietly before launch; some fail spectacularly publicly, becoming another media horror story about AI. Why does this happen? Because the current process for designing AI products and services is broken. This talk walks through a new method that has been developed over many years at Carnegie Mellon University's Human-Computer Interaction Institute. This method uses elements of user-centered design and technology capabilities to find situations where moderate technical performance, high value, and low risk combine to make successful AI projects.

Related Media

Vimeo Video

Other Resources / Information

https://hcii.cmu.edu/

Takeaways

  1. AI can be valuable while still being dumb as long as it is high value and low risk—for users and organizations.
  2. Most commercially successful AI features are simple inferences with good performance or hard inferences with moderate performance.
  3. A new approach to designing AI is possible, one that instills more cooperation between designers, PMs, data scientists, and engineers.

Speakers

Organizer

Dan Saffer, Assistant Professor Of The Practice, Carnegie Mellon University

Meta Information:

  • Event: SXSW
  • Format: Presentation
  • Track: Artificial Intelligence
  • Track 2
  • Level: Intermediate
Show me another idea

Add Comments

comments powered by Disqus

SXSW reserves the right to restrict access to or availability of comments related to PanelPicker proposals that it considers objectionable.