Like many people, I use AI tools every day. I enjoy experimenting with everything from frontier models to smaller models that I can run locally, and AI has become a significant part of both my work and personal life. I have been using these tools since the release of ChatGPT with GPT-3.5. I remember being blown away by the capabilities of this seemingly intelligent chatbot. My curiosity soon led me down a path of trying to understand more about what was going on under the hood of large language models, usually shortened to LLMs.

I’ve always taken more photographs and videos than I can realistically revisit. My iCloud Photos library has grown into a huge collection of family moments, holidays, ordinary days, and small things that seemed worth capturing at the time. Apple Photos does a very good job of storing all of that media. The problem is that storing something and remembering it are two different things. As the library grew, older photographs and videos gradually disappeared from my day-to-day view.

I’ve spent the last few years watching AI tools move from simple chat boxes to systems that can help with much larger pieces of work. At first, I thought the model was the whole story. Pick the best model, write a good prompt, and the quality of the answer would follow. That mental model worked reasonably well for asking questions or generating a short piece of text. It started to fall apart when I asked AI to do work involving files, decisions, tools, tests, and several steps. A model might suggest a useful change, but it still needed access to the right project, clear instructions, permission to act, and a way to check whether the result was actually correct.

RSS has been part of how I stay informed for years. I like choosing the sources I follow, receiving every new article in one place, and avoiding an algorithm deciding what deserves my attention. My RSS reader of choice is Miniflux. It is fast, powerful, and gives me the control I want over my feeds. It is not the most aesthetically pleasing application to read from, though, so on iOS I use ReadKit as the interface for working through my articles.

Over the last couple of years, AI tools have slowly become part of how I work, learn, troubleshoot, and organise information. I have written before about using tools like ChatGPT, Codex, Perplexity, NotebookLM, and local models with Ollama, but Hermes Agent feels slightly different. Rather than being another chat interface, Hermes Agent is closer to a personal operating layer. It can communicate through Telegram, use tools, remember useful preferences, work with files, search previous conversations, and help route different types of work into the right context.