Introducing BYOC: Bring Your Own Context In the future, we may have 10+ flagship AI models. But then comes another problem: which one should we use? What happens when you want to switch from your default AI model? Will the new model give equally relevant responses without knowing anything about you? From a consumer’s perspective, what will truly differentiate the results? Beyond cost and model capability, I believe memory or context will become essential. The provider that knows you the best will have an advantage in delivering more relevant responses. But should one AI provider own all that context? What if you could own your memory and plug it into whichever AI tool you choose? I’ve taken a small step in this direction by building The Memory Machine. Its Chrome extension lets you save useful context while browsing and retrieve relevant memories when prompting ChatGPT. I’d love to know what you think. How do you currently carry context between your browsing, research and AI conversations?

The Memory Machine

Build memory as you browse and use while prompting any LLM

Build memory as you browse and use while prompting any LLM