Meta says its smallest Muse model matches Claude Haiku 5.5 for a third of the cost. Plus a former Meta developer on fragile AI-written code, Infineon's 300 mm silicon carbide wafer, parking-app fees across 312 US cities, and a drone that passed for a bird.
- Muse Ember: Our smallest, most efficient model (HN discussion)
- AI usage generates a fragile codebase (HN discussion)
- The first 300-millimeter power silicon carbide (SiC) wafer (HN discussion)
- Parking-app convenience fees in 312 U.S. cities (HN discussion)
- Drones built to look like birds evaded detection at Air Force tech exercise (HN discussion)
Chapters:
(00:00) Intro
(00:43) 1. Muse Ember: Our smallest, most efficient model
(01:39) 1. Muse Ember: Our smallest, most efficient model: HN comments
(02:39) 2. AI usage generates a fragile codebase
(03:54) 2. AI usage generates a fragile codebase: HN comments
(05:56) 3. The first 300-millimeter power silicon carbide (SiC) wafer
(06:55) 3. The first 300-millimeter power silicon carbide (SiC) wafer: HN comments
(07:41) 4. Parking-app convenience fees in 312 U.S. cities
(08:33) 4. Parking-app convenience fees in 312 U.S. cities: HN comments
(09:24) 5. Drones built to look like birds evaded detection at Air Force tech exercise
(10:30) 5. Drones built to look like birds evaded detection at Air Force tech exercise: HN comments
(11:08) Wrap-up
More episodes from "HN Digest"
Don't miss an episode of “HN Digest” and subscribe to it in the GetPodcast app.
