Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
hci.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social platform for Human-computer Interaction researchers and practitioners. Hosted by Princeton HCI. Managed by a group of volunteers.

Administered by:

Server stats:

415
active users

hci.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.6

Daniel Buschek

Mobile email UIs added AI replies "on top". These popups hide draft & email, so users need to remember all info from the email while prompting & checking AI results. We've redesigned this with local responses, resulting in flexible workflows - both with & without AI. 👇🧵

Two screenshots with annotations. Left: Gmail showing an empty email draft with a popup on top that shows user prompt and AI reply. Right: Redesign that shows manual and AI reply snippets in small cards directly inserted into the text of the incoming email. Additional text says points out this difference in the use of space.
Mar 25, 2025, 01:10 PM·Public·Web
2boosts·0favorites
Public
Daniel Buschek

Inspired by microtasking principles, tap any sentence in the incoming email to write a local reply (or get sentence suggestions). Then connect these local replies on a usual draft screen (or let AI do just that).

Shows two screenshots of the UI. Left: incoming email with two local responses added in card-like UIs inline between the sentences of the original email. Right: AI-suggestion for a whole email based on the entered local replies. Formatted with AI changes highlighted in colour, with buttons for accept/discard below.
Public
Daniel Buschek

As our study (N=126) showed, our design (center) involves users more, also while reading. While full AI replies (bottom) are fast, our design also covers fast workflows: Users can simply skip local responses to get an AI draft, if they situationally prefer speed over control.

Shows time spent on reading/writing in the reply process as a horizontal bar chart. Three bars compare the new “local responses” design against two baselines, namely fully manual typing and full AI reply generation. Colored bar segments and line styles show that the local responses sit inbetween the baselines in terms of total time and allowed participants to spend a higher proportion of time on the reading part, since they could also add local replies while reading.
Public *
Daniel Buschek

💡Lesson learned: Rather than simply adding AI "on top" of existing UIs, (re)design for users' underlying workflows – both with & without AI. In this way, flexible UIs empower users to dynamically decide on AI involvement themselves. preprint: arxiv.org/abs/2502.06430

arXiv logo
arXiv.orgContent-Driven Local Response: Supporting Sentence-Level and Message-Level Mobile Email Replies With and Without AIMobile emailing demands efficiency in diverse situations, which motivates the use of AI. However, generated text does not always reflect how people want to respond. This challenges users with AI involvement tradeoffs not yet considered in email UIs. We address this with a new UI concept called Content-Driven Local Response (CDLR), inspired by microtasking. This allows users to insert responses into the email by selecting sentences, which additionally serves to guide AI suggestions. The concept supports combining AI for local suggestions and message-level improvements. Our user study (N=126) compared CDLR with manual typing and full reply generation. We found that CDLR supports flexible workflows with varying degrees of AI involvement, while retaining the benefits of reduced typing and errors. This work contributes a new approach to integrating AI capabilities: By redesigning the UI for workflows with and without AI, we can empower users to dynamically adjust AI involvement.
ExploreLive feeds
About