I was scrolling through a Reddit thread drafting an obituary for Alfred, a Mac launcher I have used so long my fingers move before my brain does. To the commenters, Alfred is stagnant because it isn’t releasing visual updates or shipping at the cadence of Raycast.

I worry when we assume that if a piece of software isn’t constantly overhauling its UX, it must be dying.

There is a big gap between active software and a stable business. A tool that sears meat and lasts for years does so because its technology is local and needs no external permission to work. A smart appliance, by contrast, only works as long as the servers stay on and venture capital keeps subsidizing the lights.

The Perceived Decay

On Reddit, the consensus was clear: Alfred is “in trouble” because it hasn’t pivoted, announced an AI feature, or rolled out a flashy new UI. The community seems to conflate iterative feature churn with foundational stability.

This isn’t an anti-AI argument. I use LLMs daily for coding and website fixes. But we now judge a tool’s health by how often its chrome changes rather than the structural durability of the business behind it.

The Economics of Longevity

Alfred is a local-first tool with almost no dependence on cloud instances. Its primary costs are straightforward: a small team ensuring it stays fast and compatible with macOS updates. Its survival isn’t tied to a growth-at-all-costs mandate. It costs very little to run, and the Powerpack pays the bills.

Raycast, by contrast, is a venture-backed experiment. Innovation costs money, and trying different models is fine. But shifting from a keyboard launcher into an AI hub and expanding to Windows changes the cost structure entirely. It requires constant external compute and ongoing capital. If the cash or the cloud infra disappears, the tool goes with it.

There is also a real cost to maintaining viral community hype. A vocal community can amplify features and drive adoption, but when unit economics are sound, you don’t need continuous hype cycles to survive.

Sovereignty over Scale

After years in corporate tech, I find myself drawn to software built like a hearth rather than a high-rise. I respect tools that optimize for longevity over scale, built for users who care more about reliability than endless novelty. You pay for a tool that works, not an implied promise that it will reinvent itself every six months.

Update: Right after I hit publish, Raycast rolled this out:

They are evolving into a software house, and the new product looks interesting. But just like the Arc and Dia saga showed us, shipping flashy new things does not guarantee survival, and staying focused doesn’t mean Alfred is dead.