If all you want is a way to talk to ChatGPT, you’re practically spoiled for choice — there are countless ways to chat with the bots, and seemingly countless new ones every single day. But that’s not the whole story of how we’ll interact with large language models. It better not be.
Thomas Paul Mann, CEO and cofounder of the app Raycast, has a much bigger vision for what an AI app might do. Raycast is many things: it’s an app launcher, it’s a way to search and interact with the files on your computer, it’s a note-taking app, and, yes, it’s yet another way to talk to ChatGPT and other LLMs. Because it has so much access to your data and your device, though, Raycast can use these AI models to actually do things on your behalf. Agentic AI, baby! Only this time, with even more potential — for better and for worse.
On this episode of The Vergecast, the first in our two-part series about how developers are using AI and integrating it into their products, Mann explains his big and small ideas for AI in Raycast. In the same way that so many companies are hoping to integrate their bots with browsers, in order to access all your history and preferences and Chrome-instilled muscle memory, Raycast thinks it can accomplish something very similar by replacing your Mac’s Spotlight and your PC’s Start menu. It can help you create, manage, and organize files, but it can also operate inside of any app you have installed. It can, in theory, even open up the Terminal and go wild with it. (That last one may not be advisable in most cases.)
There are, of course, huge questions raised by all this integration and access. It’s one thing for a chatbot to make mistakes in a text chat; it’s another entirely to turn a hallucinating, imperfect tool loose on your computer. AI agents, by and large, don’t really work, so why would they be more reliable when they’re working on local files and not the whole internet? And even if all this stuff does eventually work, how are we supposed to use it? Mann has some answers, and some questions of his own.
If you want to know more about everything we discuss in this episode, here are some links to get you started:
Read the full article here
