Mathieu Triay
@mathieutriay.bsky.social
Type & Interaction Bricoleur — mathieutriay.com
Really good article. We’re trying to find alternatives at BBC R&D and published some thoughts about discovery without algorithms (in this case, for music) www.bbc.co.uk/rd/articles/...
Designing for intentional discovery
Searching for music, imagine that instead of an algorithm, you had a map of the content that you could explore. That's what we set out to design with Orbit.
www.bbc.co.uk
August 6, 2025 at 6:13 PM
Really good article. We’re trying to find alternatives at BBC R&D and published some thoughts about discovery without algorithms (in this case, for music) www.bbc.co.uk/rd/articles/...
You have to display all of them because you can’t tell which one will strike, right? And it’s a different one for everyone, potentially. The process itself is great, definite inspiration for our music discovery service in BBC R&D www.bbc.co.uk/rd/articles/...
Designing for intentional discovery
Searching for music, imagine that instead of an algorithm, you had a map of the content that you could explore. That's what we set out to design with Orbit.
www.bbc.co.uk
July 30, 2025 at 7:16 PM
You have to display all of them because you can’t tell which one will strike, right? And it’s a different one for everyone, potentially. The process itself is great, definite inspiration for our music discovery service in BBC R&D www.bbc.co.uk/rd/articles/...
The future isn't just about working out where AI agents can help, but carefully working out the human impact of sharing our agency with them.
June 14, 2025 at 7:50 AM
The future isn't just about working out where AI agents can help, but carefully working out the human impact of sharing our agency with them.
This creates challenges around trust and responsibility because we're expected to have confidence in these systems while knowing they can fabricate information that sounds completely plausible.
June 14, 2025 at 7:50 AM
This creates challenges around trust and responsibility because we're expected to have confidence in these systems while knowing they can fabricate information that sounds completely plausible.
AI Agents show real promise when there's uncertainty and knowledge gaps thanks to their ability to plan and execute independently. But agency is finite so when we delegate to agents, we're not just gaining efficiency, we're also redistributing control.
June 14, 2025 at 7:50 AM
AI Agents show real promise when there's uncertainty and knowledge gaps thanks to their ability to plan and execute independently. But agency is finite so when we delegate to agents, we're not just gaining efficiency, we're also redistributing control.