Radaric
I like the ambition here. Feed readers have been around forever, but most of them just sort by date or source. The "learns your editorial taste" bit is what separates Radaric from a glorified RSS aggregator. It's betting that your reading habits are a signal worth mining, and honestly, they are. The founders who read deeply tend to spot trends early, so a tool that sharpens that instinct has real appeal.
What I'd want to know is how it actually learns. Does it watch what you open, what you skip, how long you linger? Does it adapt to your mood, or is it locked into one static profile? The hard part of taste is that it shifts. You might want deep dives on Monday and quick takes on Friday. If Radaric can handle that nuance, it's onto something. If it just builds a generic preference model, it'll feel like every other recommendation engine.
I'm cautiously optimistic. The feed reader space is crowded, but nobody has fully cracked taste-based curation yet. Radaric has a shot if it keeps the learning loop tight and doesn't drown users in settings. I'll be watching to see how it handles the first few weeks of usage.