Assumption Testing P2 – SnapEdit (Team 12)

By Armita Hosseini & Benjamin Klieger

Notes on team discussion:

  • When AI works, it has a significant time advantage. We used a simulation of AI helping on specific tasks, it is possible our actual AI may not be as helpful. We should keep that risk in mind. 
  • AI can help for simple design tasks, i.e. make a square in the center of the page, or add an image of this thing.
  • Time advantage from 6:00 to 1:30 is very significant
  • Weird shapes and visual problems push designer to delete the entire graphic when it does not work
  • We should detect when the AI does not work well, i.e. user is deleting or reversing decisions, and learn from those cases
  • We should give control to user and frame AI as something that can help but will not always help

Link to Armita’s interview transcript: https://docs.google.com/document/d/1dzranRJZw4ht7Nv3iBe-czyB4K9hwd1Pf1LxCk5yImM/edit?usp=sharing

Link to Ben’s interview transcript: https://docs.google.com/document/d/1FOlZBlOi-rA0Hj61WjrGYH_sypXDmDwc9tZ1pgVdeAY/edit?usp=sharing

Avatar

About the author