User Defenders – UX Design & Personal Growth · Jason Ogle

The State of UX in AI with Josh Clark

·1 hr 54 min·5 clips
Josh Clark exposes how Google's algorithms once confirmed harmful stereotypes like 'women are evil' in search results.
A listener named Sherry Benko sparks the topic. Jason looks at the state of UX in AI as a question of judgment: when should automation pause, admit it is unsure, and pull a person back into the loop? The episode is thoughtful without getting misty about machines. AI is fast, useful, and still bad at knowing when to stop. Alexa gives the lesson some teeth. The household voice device understands Jason and Veronica about 90 percent of the time, then fails Liza completely. That miss lands harder than a normal software error. A request to turn on the lamps somehow becomes Christian children's music, and a convenience feature turns into family irritation. Jason catches the bigger point quickly. A talking interface may only be running an algorithm, but users do not always feel it that way. When the machine sounds social, the failure can feel social too. Intent is the other problem. Speech recognition can capture the words and still miss the meaning, leaving the user stranded with a weirdly confident device. Jason keeps bringing the answer back to humility: show uncertainty, make the handoff clear, and invite human judgment before the experience gets awkward. The superhero frame gives the craft some charm. The Alexa story keeps it grounded. The close thanks listeners for sending topics and points toward more reflective monologues, including one on imposter syndrome. Fight on, my friends.

As heard by us

A case for keeping human judgment in AI interfaces.

The episode argues that human judgment still matters most when an algorithm falters or loses confidence. Its strongest examples come from a voice assistant that can seem helpful to one person, exclusive to another, and wrong even when it hears the words correctly.

Read the full review in PlayNext →

Why you'd press play

If AI UX feels slippery, this episode gives you a steadier way to judge where it breaks in practice.

Read the full recommendation in PlayNext →
Listen to the show on